OPINION
SAYLOR, Justice.
Appeal was granted in this case to address the standards of proof applicable in real estate tax assessment appeals.
Appellants, Robert C. Green and Judith A. Green (“Taxpayers”), are the owners of a 6,344-square-foot, single-family residence situated on 1.8 acres in Schuylkill County. The residence is a two-story home, built in 1987, with, inter alia, a stone and brick exterior, a shingled roof, four bedrooms, two full baths, several partial baths, a four-car garage, and a basement that is 60 percent finished. In 1997, pursuant to a [189]*189countywide reassessment, Taxpayers’ home was assessed based upon a fair market value of $612,580. Taxpayers appealed the assessment to the Schuylkill County Board of Assessment Appeals (the “Board”), which denied the appeal following a hearing. Taxpayers then appealed the Board’s decision to the Court of Common Pleas of Schuylkill County.
At the de novo hearing before the trial court, the Board introduced the official assessment record into evidence and then rested. To counter the Board’s evidence, Taxpayers presented the testimony of Anthony Matsell, a real estate appraisal expert, who offered the opinion that the fair market value of Taxpayers’ property, as of January 1, 1996, was $360,000, The path by which he reached this conclusion was explored in depth on direct and cross-examination.
At the outset, Mr. Matsell explained that he had considered the three methods of valuation specified in Section 402 of the General County Assessment Law, 72 P.S. § 5020-402, namely, the cost approach method, the comparable sales method, and the income approach.1 According to Mr. Matsell, the cost approach method was not a reliable indicator of the property’s fair market value because Taxpayers’ property, given its geographical location, was over-improved. Consequently, he reasoned, “people in that market area won’t pay the amount it would cost to reproduce this house,” which he estimated at $500,614, since they would not be able to recoup that amount in a later sale.2 More generally, the expert noted, he had [190]*190never seen a buyer rely on the cost approach in determining how much to pay for a residential property. As for the income approach, Mr. Matsell testified that this method of valuation was also inappropriate for Taxpayers’ property, a personal residence that does not generate any income.
Accordingly, Mr. Matsell explained, he relied primarily upon the comparable sales method — “comparing your property against similar properties that had sold on the market [recently]” — in establishing a fair market value for the property. Owing to the over-improved nature of Taxpayers’ property, Mr. Matsell had a “very, very hard time finding anything like it.” Homes in Taxpayers’ neighborhood ranged in value from $150,000 to $300,000, with an average of $180,000, and most were - 3,000 to 3,500 square feet in size; countywide, Mr. Matsell “found no properties that [had] sold in excess of ... $400-450,000.”3
Accordingly, Mr. Matsell chose, as the best available “com-parables,” the following three properties:
1. A 3,900-square-foot residence situated on 1.76 acres, located within a block of Taxpayers’ residence, with an actual sale price in 1995 of $335,000 and an adjusted sale price of $360,000;4
[191]*1912. A 3,000-square-foot residence situated on 1.285 acres, located within a block of Taxpayers’ residence, with an actual sale price in 1995 of $252,500 and an adjusted sale price of $351,000; and
3. A 3,400-square-foot residence situated on .8 acre, located 1.5 miles from Taxpayers’ property, with an actual sale price in 1993 of $415,000, and an adjusted sale price of $444,000.
Although Comparable # 3 had the highest sale price, Mr. Matsell relied most heavily on Comparables # 1 and # 2 in valuing the subject property because, as he explained, they were “only a block away from the subject and more within [its] market area”; in addition, they had sold more recently than Comparable # 3. Using the comparable sales method, Mr. Matsell calculated the fair market value of Taxpayers’ home as $360,000, adding that “nobody’s paid anything much more than I have appraised the subject property at.” After Mr. Matsell concluded his testimony, Taxpayers rested; the Board did not present any testimony or evidence in rebuttal.
The trial court accepted the valuation of $360,000 reached by Taxpayers’ expert. In its opinion, the trial court explained that it had found some aspects of the expert’s testimony (for example, his conclusion that Taxpayers’ property was super-improved) to be credible.
On the other hand[, the trial court continued,] we did find portions of Maxell's testimony suspect. For example, while he opined that " was impossible to find a comparable sale in excess of the value he had placed upon this particular property in the market area, comparable number 3 on his market report showed an unadjusted sales value of $415,000 [192]*192for a similar sized parcel with less improvements. After adjustments, comparable number 3 had an adjusted sales price of $444,000, which was in contradiction to his testimony as to the maximum fair market value that the subject property could demand within the market area. Furthermore, his testimony that he could find no properties that sold in excess of $400,000 to $450,000 was contradicted by his own testimony under cross-examination. Nor did we find compelling Mr. Matsell’s reasons for relying more so on his number 1 and number 2 comparables, rather than comparable number 3.
Trial Court Opinion at 4 (citations omitted).5 Ordinarily, the trial court explained, it would have dealt with such a situation by applying the settled principle that the factfinder is free to believe all, part, or none of the evidence.
The property here was a residential parcel within a relatively narrow geographic area to which most finders of fact, including the trial court, could reasonably assess some valuation. In so doing we normally would consider the cost approach, as limited by testimony as to overimprovement, as well as a revealing comparable which would have set a fair market value somewhat above that of the expert witness.
Id. at 4-5. In the trial court’s view, however, the Commonwealth Court’s decision in 841 Associates v. Board of Revision of Taxes, 674 A.2d 1209 (Pa.Cmwlth.1996), meant that such latitude was no longer available to it. In 841 Associates, the Commonwealth Court had held that “[i]f the expert witness is credible, then the unrebutted evidence of valuation given by the witness must be accepted by the trial court. It is the expert’s valuation that the trial court must accept!,] not just a piece of her testimony.” Id. at. 1214 (citations omitted). Interpreting 841 Associates in the context of the case sub judice, the trial court concluded that where, as here, the taxpayer’s expert “was not without credibility” and the Board had failed [193]*193to offer expert testimony of its own, the valuation reached by the taxpayer’s expert had to be accepted.
The Board appealed to the Commonwealth Court, arguing that the trial court had erred in interpreting 841 Associates. A divided court en banc agreed. Rejecting the possibility that the trial court had “painted 841 Associates with too broad a brush,” the majority explained that
the plain language of our opinion in that case supports the trial court’s approach.... Followed literally, 841 Associates imposes upon the trial court the Hobson’s Choice that it must either credit or discredit an unrebutted expert’s testimony in its entirety, and mandates that the trial court must accept an expert’s final valuation even where it has found incredible one or more component(s) of the testimony necessary to calculate that value.
Green v. Schuylkill County Bd. of Assessment Appeals, 730 A.2d 1017, 1020-21 (Pa.Cmwlth.1999) (en banc) (emphasis in original). This was incorrect, the majority concluded: under controlling precedent, particularly this Court’s decision in Westinghouse Elec. Corp. v. Board of Property Assessment of Allegheny County, 539 Pa. 453, 652 A.2d 1306 (1995), “the fact-finder is not constrained to accept the ultimate opinion of an expert merely because the witness is unrebutted and has provided some credible testimony.” Id. at 1021 (footnote omitted; emphasis in original). To the contrary,
the only constraint upon the fact-finder’s determination of fair market value is that it must be supported by expert testimony found to be credible, and may not be based, even in part, upon the assessment record once any expert testimony is credited.
Id. at 1022 (citation deleted; emphasis in original). Accordingly, the majority overruled 841 Associates and remanded the case sub judice to the trial court with instructions “[to] determine the weight to accord the expert’s testimony and, based upon the evidence of record found to be credible, determine the fair market value for the property at issue.” Id. at 1022.
[194]*194In a dissenting opinion, Judge Pellegrini, joined by Judge Flaherty, expressed disagreement with the majority’s conclusion “that the trial court can pick and choose among the factors used by the expert, reweigh them, and then arrive at its own independent opinion of value.” Id. at 1022 (Pellegrini, J., dissenting). In fact, asserted the dissent, settled precedent precludes the trial court from finding a fair market value different from that found by an expert whose testimony is unrebutted. See id. at 1023.
We allowed appeal to determine whether it is the Commonwealth Court’s decision in 84-1 Associates, or its subsequent decision overruling the same, which comports with the principles underlying this Court’s tax assessment jurisprudence. Resolving this issue will require us to examine and address the role of the trial court in tax assessment appeals.
We begin our analysis' by delineating the procedural framework within which the trial court performs its function. The assessment of real estate taxes is governed by the General County Assessment Law, 72 P.S. §§ 5020-1-5020-602, and by the assessment law applicable to the particular class of county — in the case of Schuykill County, a fourth-class county, the Fourth to Eighth Class County Assessment Law, 72 P.S. §§ 5453.101-5453.706. Both statutes direct the assessor to rate and value all objects of taxation “according to the actual value thereof....” 72 P.S. § 5020-402(a); 72 P.S. § 5453.602(a).6 A taxpayer who wishes to challenge the valuation, or any other aspect of the assessment, may appeal to the Board of Revision of Taxes, see 72 P.S. §§ 5020-511, 5453.701, 5453.702, and, if such appeal proves unsuccessful, to the court of common pleas, see 72 P.S. §§ 5020-518.1, 5453.704.
[195]*195In an assessment appeal, the matter before the trial court is heard de novo, and the order of proof is well settled. Deitch Co. v. Board of Property Assessment, 417 Pa. 213, 221, 209 A.2d 397, 402 (1965).
The procedure requires that the taxing authority first present its assessment record into evidence. Such presentation makes out a prima facie case for the validity of the assessment in the sense that it fixes the time when the burden of coming forward with evidence shifts to the taxpayer. If the taxpayer fails to respond with credible, relevant evidence, then the taxing body prevails. But once the taxpayer produces sufficient proof to overcome its initially allotted status, the prima facie significance of the Board’s assessment figure has served its procedural purpose, and its value as an evidentiary devise is ended. Thereafter, such record, of itself, loses the weight previously accorded to it and may not then influence the court’s determination of the assessment’s correctness.
[T]he taxpayer still carries the burden of persuading the court of the merits of his appeal, but that burden is not increased by the presence of the assessment record in evidence.
Of course, the taxing authority always has the right to rebut the owner’s evidence and in such a case the weight to be given to all the evidence is always for the court to determine. The taxing authority cannot, however, rely solely on its assessment record in the face of countervailing evidence unless it is willing to run the risk of having the owner’s proof believed by the court.
Id. at 221-22, 209 A.2d at 402 (citations and footnote omitted).
The trial court’s statutory mandate, as established in the General County Assessment Law, is to hear the evidence and to “make such orders and decrees ... as ... may seem just and equitable....” 72 P.S. § 5020-518.1(a). The Fourth to Eighth Class County Assessment Law includes a more specific direction to the trial court to determine, inter alia, the market value of the subject property. See 72 P.S. § 5453.704(b). In essence, “[t]he Legislature has confided to the Court of Com[196]*196mon Pleas the duties of fact finder where there has been an appeal from an assessment for taxes.” Appeal of Edmonds, 314 Pa. 382, 384, 172 A. 103, 104 (1934); accord Appeal of Park Drive Manor, 380 Pa. 134, 136, 110 A.2d 392, 394 (1955); In re Lehigh & Wilkes-Barre Coal Co., 300 Pa. 564, 566-67, 151 A. 359, 359 (1930).
This does not mean that the trial court becomes an assessor, see Appeal of Rieck Ice Cream Co., 417 Pa. 249, 256, 209 A.2d 383, 387 (1965), or an appraiser, see Appeal of F.W. Woolworth Co., 426 Pa. 583, 586, 235 A.2d 793, 795 (1967). Rather, in assessment cases, as in others, the trial court must make its determination on the basis of the evidence put before it. See Woolworth, 426 Pa. at 586, 235 A.2d at 795; Park Drive Manor, 380 Pa. at 137, 110 A.2d at 394; Appeal of Kaemmerling, 282 Pa. 78, 82, 127 A. 439, 441 (1925); In re Kemble’s Estate, 280 Pa. 441, 447, 124 A. 694, 696 (1924); Appeal of Pennsylvania Stave Co., 236 Pa. 97, 103, 84 A. 761, 763 (1912). The credibility and weight of such evidence is for the trial court to determine. See Rieck, 417 Pa. at 254, 209 A.2d at 386; Glen Alden Coal Co. v. Schuylkill County Comm’rs, 345 Pa. 159, 169, 27 A.2d 239, 244 (1942). Thus, as this Court has recently observed, “[t]he duty of the trial court in hearing a tax assessment appeal de novo is to independently determine the fair market value of the parcel on the basis of the competent, credible and relevant evidence presented by the parties.” Westinghouse, 539 Pa. at 463, 652 A.2d at 1311 (quoting In re Appeal of Jostens, Inc., 97 Pa.Cmwlth. 106, 113-14, 508 A.2d 1319, 1323 (1986) (citations omitted)).
Our review in tax assessment matters is limited to determining whether the trial court abused its discretion, committed an error of law, or reached a decision not supported by substantial evidence. See Westinghouse, 539 Pa. at 459, 652 A.2d at 1309. While the weight of the evidence is before the appellate court for review, see Chatfield v. Board of Revision of Taxes, 346 Pa. 159, 161, 29 A.2d 685, 686 (1943); In re Hudson Coal Co., 327 Pa. 247, 250, 193 A. 8, 10 (1937), the trial court’s findings of fact are entitled to great weight [197]*197and will be reversed only for clear error, see Park Drive Manor, 380 Pa. at 186-37, 110 A.2d at 394; Appeal of Carnegie, 357 Pa. 138, 142, 53 A.2d 425, 427 (1947); Chatfield, 346 Pa. at 161, 29 A.2d at 686; Appeal of Westbury Apts., Inc., 314 Pa. 130, 131, 170 A. 267, 267 (1934).
Viewing the trial court’s role as factfinder in conjunction with the applicable order of proof, it is evident that where, as here, the taxpayer offers expert testimony challenging the official assessment and the taxing authority offers no evidence in rebuttal, several factual scenarios are possible:
First, the trial court could conclude that the expert’s testimony is not worthy of belief and, therefore, that the taxpayer has failed to overcome the authority’s prima facie case. See, e.g., Appeals of Mathies Coal Co. and Consolidated Coal Co., 435 Pa. 129, 138, 255 A.2d 906, 910 (1969); Carnegie, 357 Pa. at 143, 53 A.2d at 427.
Second, the court could find that the expert’s testimony is competent and credible in all respects. In that case, pursuant to settled law, the expert’s opinion as to fair market value must be accepted. See Woolworth, 426 Pa. at 586, 235 A.2d at 795; Deitch, 417 Pa. at 222, 209 A.2d at 402; Rieck, 417 Pa. at 254, 209 A.2d at 386; Kaemmerling, 282 Pa. at 83, 127 A. at 441; Kemble’s Estate, 280 Pa. at 447, 124 A. at 696; Shannopin Coal Co. v. Greene County, 280 Pa. 4, 8-9, 124 A. 266, 267 (1924). As this Court stated more than eighty years ago,
[i]f the testimony as to the value of any particular tract of land is uncontradicted, and is worthy of belief, that valuation should be conclusive; if the testimony be conflicting, the court must use its best judgment in determining from the weight of it what is a just valuation.
Lehigh Valley Coal Co. v. Northumberland County Comm’rs, 250 Pa. 515, 525, 95 A. 712, 714 (1915).
Third, the court could accept the taxpayer’s expert’s testimony as wholly credible, but recognize the need to correct or modify the ultimate valuation figure due to a simple error, for example, a mathematical error that can be corrected utilizing principles grounded in common experience.
[198]*198Fourth, the trial court could find that the expert’s testimony is sufficient to overcome the authority’s prima facie case, but not credible in all respects pertaining to the valuation ultimately reached by the expert. The difficulties posed by such a scenario are obvious: the trial court would have to determine the fair market value without adopting the estimate offered by the expert, but also without taking upon itself the role of appraiser or assessor. The Commonwealth Court addressed this factual situation in 841 Associates.
The de novo proceedings in that case 'nvolved a dispute over the fair market value of a Philadelphia office building. The taxing authority, which had determined that the property had a fair market value of $43.2 million, introduced its assessment record into evidence and then rested. In response, the taxpayer presented the testimony of an expert witness, Teresa Hoberg. Ms. Hoberg testified that she had valued the property using the three standard valuation methods (comparable sales, cost, and income approach). The comparable sales method produced a valuation of $24 million. The cost approach, according to Ms. Hoberg, was inappropriate given the age of the building. To arrive at a valuation via the income approach, Ms. Hoberg performed two computations: one based on the stabilized income approach, which produced a valuation of $25.5 million; and another based on a discounted cash flow analysis, which produced a valuation of $25 million. Ms. Hoberg explained that she combined thése results for an income-approach valuation of $25 million, and then correlated this figure with the comparable-sales valuation, for a final fair market value of $25 million. See id. at 1211-12.
On cross-examination, counsel for the taxing authority asked Ms. Hoberg to perform the discounted cash flow analysis utilizing a different capitalization rate (13.28 percent). Ms. Hoberg did so, and arrived at a fair market value of $32.7 million. She explained, however, that she would not choose to utilize a rate of 13.28 percent, as it was inappropriate and violated industry standards. See id. at 1212.
[199]*199The taxing authority provided no rebuttal evidence.7 The trial court found that the fair market value of the property was $32.5 million, explaining that it credited “some of Ho-berg’s testimony,” id. at 1212, specifically, that in which she utilized the 13.28 percent capitalization rate to arrive at a valuation of $32.7 million. See id.
On appeal, a divided panel of the Commonwealth Court reversed. The issue presented, according to the majority, was a purely legal one: “whether a trial court may ... accept only part of the Taxpayer’s expert testimony, but not the valuation, when the taxing authority fails to present rebuttal evidence.” Id. at 1211. Curiously, given its phrasing of the issue, the majority grounded its analysis on the assumption that “the trial court found [that] Hoberg’s testimony [was] competent and relevant and that she was a credible witness....” Id. at 1214 (emphasis added). That being the case, the majority reasoned, the trial court had erred in failing to “accept her expert valuation of the property, instead, ... adopting one number from part of one method of determining fair market value.” Id. “Where the taxpayer’s testimony is relevant, credible and unrebutted,” the majority emphasized, “it must be given due weight and cannot be ignored by the court. It must necessarily be accepted.” Id. at 1213 (quoting Deitch, 417 Pa. at 222, 209 A.2d at 402). Accordingly, the majority concluded that, “[bjecause Hoberg’s testimony was the only competent, credible and unrebutted evidence of valuation,” id. at 1214, the trial court erred in not accepting her valuation of $25 million. See id.
In a vigorous dissent, Judge Narick asserted that, in view of the trial court’s limited finding of credibility, the majority’s reliance on Deitch for the effect of “relevant, credible and unrebutted” testimony was “completely puzzling.” 841 Associates, 674 A.2d at 1215 (Narick, J., dissenting). What the trial court had found, the dissent emphasized, was that Ms. [200]*200Hoberg was “only partially credible,” id., an entirely appropriate finding given the trial court’s ability as factfinder to believe all, part, or none of an expert witness’s testimony. See id. The dissent asserted that the majority’s failure to acknowledge the precise scope of the trial court’s credibility determination was a blatant violation of this Court’s precedent regarding the trial court’s role as the sole arbiter of credibility. See id. In the dissent’s view, the trial court’s credibility determination should be upheld and its determination of fair market value, which rested on that credibility determination, should be affirmed. See id. at 1216.
As the dissent suggested, the majority’s analysis is somewhat paradoxical: it acknowledges that the trial court found the taxpayer’s expert to be “only partially credible,” but refers throughout the opinion to the conclusive effect of “credible,” unrebutted testimony. The apparent confusion is clarified by the majority and dissenting opinions in the present case, which, like 811 Associates, involves a scenario in which the taxpayer’s expert’s testimony is unrebutted but not, in the trial court’s view, entirely credible.
The unavoidable implication of 811 Associates, notes the majority here, is that where the testimony of a taxpayer’s expert is unrebutted, the trial court “must either credit or discredit [such] testimony in its entirety.” Green, 730 A.2d at 1021 (emphasis in original). The dissent, written by the author of the majority opinion in 811 Associates, confirms the majority’s interpretation. The trial court’s role in a case involving unrebutted expert testimony, the dissent reasoned, differs from its role in a case such as Westinghouse, where both parties offer experts with conflicting opinions of the fair market value of the subject property. See Green, 730 A.2d at 1024-25 (Pellegrini, J., dissenting). In the latter type of case, according to the dissent, it is reasonable for the trial court, if it finds both experts credible, to “select[ ] values within the extremes represented by the credible evidence.” Id. at 1025 (quoting Westinghouse, 652 A.2d at 1316 (Flaherty, C.J., concurring and dissenting)). In the former type of case, however,
[201]*201[lacking] two opinions of value, there is nothing that the factfinder can use to triangulate its position. This casts the factfinder adrift in search of a value, causing it to become something that our Supreme Court says it cannot be — an expert appraiser.
Id. Addressing the specific conclusions that the trial court suggested it would reach if not foreclosed by 841 Associates, the dissent asked:
[U]pon which evidence would the factfinder rely to determine that there are better comparable sales? Where would the factfinder get this evidence? Where would the factfinder get the expertise to determine that the real estate expert relied more heavily on some comparables than others?
Id. at 1023 (footnote omitted). Relying principally on this Court’s decisions in Woolworth and McKnight, the dissent concluded that, in a case such as the present one, “where the taxpayer’s real estate expert’s testimony [is] unrebutted, the trial court [has] two options: accept the valuation or find the expert not credible.” Green, 730 A.2d at 1024 (Pellegrini, J., dissenting). In the dissent’s view, therefore, a fourth scenario — crediting some but not all of the expert’s testimony — is not possible.
Taxpayers contend that it is the dissent’s reasoning, not the majority’s, that comports with the well-settled standard of proof established by this Court’s precedents. According to Taxpayers, the majority has adopted a new standard of proof, pursuant to which “the trial court may now disregard competent, credible and unrebutted evidence of market value provided by the landowner’s expert and set its own value for the landowner’s property, despite the fact that any such finding of value by the trial court would not be based on the evidence presented at trial.” They ask that, in addition to vacating the Commonwealth Court’s order and reinstating that of the trial court, we expressly affirm the decision of the majority in 841 Associates.
[202]*202Taxpayers misapprehend both this Court’s holdings and the role that the Commonwealth Court majority would allow the trial court to play in an assessment appeal. As noted earlier, this Court has held on numerous occasions that where a taxpayer’s expert’s testimony is both unrebutted and credible, the valuation testified to by the expert must be accepted. In those decisions, however, there was no indication that the expert was anything but entirely credible. See, e.g., Rieck, 417 Pa. at 255, 209 A.2d at 386 (noting, after requiring the trial court to accept the taxpayer’s unrebutted evidence of market value, the absence of any “finding or even suggestion that [the taxpayer’s] witnesses are not credible” (quoting Kemble’s Estate, 280 Pa. at 447, 124 A. at 696)).
Woolworth and McKnight, the decisions relied upon by the dissent, are not to the contrary. The taxpayers in Woolworth challenged the assessment of their property by calling as an expert witness a real estate broker, who testified that the market value of the property was $975,000. Although the Board of Property Assessment called several rebuttal witnesses, their testimony did not address the property’s fair market value. The trial court commented favorably on the taxpayers’ expert’s testimony, noting that he possessed a thorough knowledge of real estate in the relevant area and had made a comprehensive study of the subject property. Nevertheless, the court determined without explanation that the fair market value of the property in question was not $975,000, as the expert had found, but $1,313,700. On appeal, this Court reversed, explaining that
[n]o fair reading of the record will support the Board’s contention that the court below basically rejected the market value testimony of taxpayers’ expert witness. In fact, the opposite is true; and the Board proceeded at its peril in not producing any testimony whatsoever to refute this expert opinion. In such a situation we have clearly held that the taxpayers’ evidence must be accepted. The court is not an expert appraiser; it can act only upon credible and competent evidence presented to it. Having found such evidence before it, the lower court here ... could only [203]*203consider the assessment of the Board to have been rebutted and must determine the fair market value to be $975,000.
Id. at 586, 235 A.2d at 795 (citations omitted; emphasis added).
The matter at issue in McKnight was not fair market value, but uniformity (that is, whether properties in a given taxing district are assessed at the same ratio of assessed to market value). The subject property, a shopping center, had been assessed at $486,970. During the de novo proceeding in the trial court, the property owner offered an expert witness who testified that the subject property had been assessed at 88.5% of its market value, while two other shopping centers, considered by the witness to be comparable, had been assessed at only 57% and 76% of their market values. See id. at 237, 209 A.2d at 391. The trial court dismissed the appeal after finding that the taxpayer’s expert’s testimony as to lack of uniformity “was not competent.” Id. at 238, 209 A.2d at 391.8
In the view of this Court, reviewing the matter on appeal, “it [was] clear that the [trial] court rejected the testimony of [the expert] witness on the theory that it could not be considered credible.” Id. The trial court’s finding in this regard was understood to derive from, inter alia, the assumption that a shopping center having only one-third the acreage and one-half the square footage of the subject property, as was the case with one of the comparables, could not, contrary to the expert’s testimony, have a market value equal to 80% of the subject property’s market value. See id. That assumption was erroneous, the Court explained: the trial court had failed to recognize that size is but one of many factors that determine a property’s market value. See id. at 239, 209 A.2d at 392 (quoting Park Drive Manor, 380 Pa. at 136, 110 A.2d at 394). Thus, it was quite possible that a smaller shopping center could have a fair market value equal to 80% of the [204]*204larger’s. See id. On this point, the Court made the following observation:
While we do not view the testimony of [the taxpayer’s] witness as completely satisfactory in explaining fully the factors applicable here, this testimony cannot be disregarded unless the lower court in the exercise of its discretion finds it incredible. No such finding is justified here....
Id. at 240, 209 A.2d at 392. Accordingly, the case was remanded to the trial court for further proceedings. See id. at 243, 209 A.2d at 394.
With the statement just quoted, this Court did not intend to say that expert testimony which is credible only in part must nevertheless be accepted in its entirety. Instead, the focus of the observation is on the substantive credibility of the testimony: because the testimony did not state an impossibility, the trial court was not justified in dismissing it as incredible.
Also relevant is this Court’s discussion of the uniformity issue in Westinghouse, 539 Pa. at 465-70, 652 A.2d at 1312-15. There, the taxing authorities made out a prima facie case that the common level ratio for the three-year period at issue was set at 50%. See id. at 465, 652 A.2d at 1312. The taxpayer was able to rebut the authorities’ prima facie case with an expert witness who testified that, based on a statistical study, the ratios for the three years in question varied from 24.38 percent to 26.04 percent. See id. at 466, 652 A.2d at 1313. The taxing authorities responded with a statistical expert who criticized the methodology of the taxpayer’s study, but did not specify alternative ratios. See id. at 458, 466, 652 A.2d at 1309, 1313. The trial court did not accept the ratios determined by the study, instead stating without explanation that the ratio for the three-year period in question was 39%. See id. at 467, 652 A.2d at 1314.
On appeal, this Court concluded that the trial court had erred by roughly “splitting the difference” between the ratio calculated by the taxing authorities and the values calculated by the taxpayer’s expert. See id. at 467-68, 652 A.2d at 1314. Such a determination, this Court explained, was arbitrary and [205]*205unsupported by the evidence. See id. at 469, 652 A.2d at 1314. Significantly, however, even though this Court had previously recognized as valid the methodology used in the taxpayer’s study, see id. at 466, 652 A.2d at 1313 (citing Appeal of Massachusetts Mut. Life Ins. Co., 426 Pa. 566, 571, 235 A.2d 790, 792 (1967)), the Court rejected the taxpayer’s argument that, because the taxing authorities had failed to counter its expert’s testimony, the trial court had erred in not accepting the ratios found by the expert.
In the instant case, the testimony of the [taxpayer’s] expert was received by the trial court as competent, credible and unrebutted evidence. That evidence, thus, must be accepted and given due weight. However, the weight that is due is for the trial court, to determine. We will not, upon this record, insist that the trial court find the common level ratio for the years in question to be what [the taxpayer’s] expert said it was. Such a course would be tantamount to this Court’s deciding the common level ratio for the relevant years. Accordingly, this question must be remanded to the trial court for resolution.
Id. at 470, 652 A.2d at 1315.9
In the case sub judice, the Commonwealth Court did not err in reaching the same result with respect to the trial court’s determination of fair market value. To hold otherwise would fail to comport with the understanding of the assessment process that is embodied in the decisional law. This Court has recognized that “[t]axation is a practical, and not a scientific, problem.” Glen Alden, 345 Pa. at 166, 27 A.2d at 243. Determining the lair market value of a property, therefore, is often “not a matter of exact science or capable of mathematical accuracy....” Hammermill Paper Co. v. City of Erie, 372 Pa. 85, 97, 92 A.2d 422, 428 (1952). To preclude the trial court in a single-expert case from determining that the [206]*206appropriate valuation lies somewhere between the official valuation and the inevitably lesser valuation offered by the taxpayer’s expert is to attribute to the expert’s valuation a definitiveness that it does not automatically possess, and to limit unnecessarily the trial court’s ability to perform its legislatively designated role as factfinder. Cf. Semasek v. Semasek, 509 Pa. 282, 288, 502 A.2d 109, 112 (1985) (observing, in an equitable distribution case, that the trial court as factfinder “need not accept even the uncontradicted opinion of a valuation expert”); Avins v. Commonwealth, 379 Pa. 202, 206, 208, 108 A.2d 788, 790-91 (1954) (explaining, in a condemnation case, that expert testimony is not proof of a fact, and that the trial court may utilize common sense in evaluating the reasonableness of such testimony).
Nevertheless, the concerns expressed by the dissent regarding the sources of evidence and the nature of the trial court’s expertise should not be dismissed out of hand, as the scope of the trial court’s factfinding authority cannot be unlimited if the court is to avoid the role of expert appraiser. The final step in our analysis, therefore, is to derive, from the principles applicable to tax assessment generally, the parameters governing the trial court’s determination of fair market value in a single-expert case.
First, and most fundamental, is the requirement that the trial court base its findings on the evidence of record. “In tax cases, [as in] all others, courts must be guided by the evidence in determining what are proper valuations.” Pennsylvania Stave, 236 Pa. at 103, 84 A. at 763. It follows from this requirement that, if the trial court is to depart from the only expert valuation contained in the record, the expert must have explained the considerations on which such valuation was based in sufficient detail that the trial court is able to evaluate the reasonableness of those considerations.10
[207]*207As an illustration, we note the comment of the trial court in the present ease that, if not bound by the constraints of 841 Associates, it would consider, inter alia, “a revealing comparable which would have set a fair market value somewhat above that of the expert witness.” If, with this comment, the trial court was expressing its intent to go beyond the evidence of record and search sua sponte for other properties that could serve as comparables, the dissent’s misgivings would be justified, as such a course of action would be improper. The likelier explanation for the comment is that the trial court was expressing the view that Comparable # 3 deserved greater weight than was given to it by Taxpayers’ expert. Comparable # 3, as noted earlier, had the highest adjusted fair market value of the three comparables, but was discounted by the expert because it was farther away from the subject property and had been sold two years earlier than the other comparables. The weight to be given such factors is, in our view, for the trial court to determine in accordance with such principles as it may derive from the record.
[208]*208Second, “[i]n making a determination in a tax assessment appeal, the trial court must state the basis and reasons for its decision.” Westinghouse, 539 Pa. at 464, 652 A.2d at 1312. This requirement is as significant to a single-expert case as to a case with competing experts, such as Westinghouse; indeed, it may be more so. Where the trial court “is presented with conflicting testimony by equally credible experts,” id. at 464, 652 A.2d at 1312, it is appropriate for the court to conclude that the fair market value of the subject property is “somewhere between the values presented by the parties.” Id.; see also Westinghouse, 539 Pa. at 472, 652 A.2d at 1316 (Flaherty, C.J., concurring and dissenting) (observing that “there is nothing arbitrary about setting the values near the midpoints of the range”). Where, however, a single expert testifies as to fair market value, and the trial court’s valuation departs from that figure, there is no basis for assuming that the trial court’s valuation is, ipso facto, appropriate. Accordingly, the trial court’s reasoning must be stated on the record so that the reviewing court may determine if the trial court’s departure from the expert’s valuation is warranted.
Although it would be premature to approve or disapprove the limited reasoning which the trial court, constrained by 8kl Associates, was able to provide, several, observations may nevertheless prove useful. First, the particular examples cited by the trial court as underlying its suspicion that Mr. Matsell’s testimony was internally inconsistent appear to be flawed. Contrary to the trial court’s assertion, Mr. Matsell did not testify that “it was impossible to find a comparable sale in excess of the value he had placed upon this particular property.” Rather, he observed that “nobody’s paid anything much more than I have appraised the subject property at”; arguably this was a fair statement, as he valued the subject property at $360,000, and the highest actual sale price among the comparable properties was $415,000. The comparably sized residences that had recently been built could not be utilized as comparables, Mr. Matsell explained, because they had not yet sold. See supra note 3. Nor did Mr. Matsell [209]*209contradict Ms own statement “that he could find no properties that sold in excess of $400,000 to $450,000”; even the highest adjusted sale price among the comparables ($444,000) was within this range.11
While Mr. Matsell’s testimony may not be “suspect” for the reasons perceived by the trial court, however, the trial court’s inclination to “set a fair market value somewhat above that of the expert witness” is not necessarily insupportable. We note, for example, that Comparable # 3 had an adjusted sale price of $444,000, the highest among the comparables, even though it was one acre less than the subject property. According to Mr. Matsell, it was not necessary to adjust the comparable’s sale price to reflect the difference in acreage. However, when asked on cross-examination whether a purchaser would not have paid more for the additional acre, he replied that “in the real world they may have.” In addition, while the comparably sized residences may have been recently constructed and therefore lack sales data, the mere fact of their construction arguably suggests that the expert’s estimation of the market for such properties (“people in that market area won’t pay the amount it would cost to reproduce this house”) is not universally shared.
These observations are in no way intended to control the determination reached by the trial court on remand. Rather, they are intended to address the dissent’s concern that the majority has “east[ ] the factfinder adrift in search of a value.” [210]*210Green, 730 A.2d at 1025 (Pellegrini, J., dissenting). To the contrary, while acknowledging that tax assessment is an inexact science, this scenario, like every other, requires that the factfinder’s valuation be tied firmly to the evidence of record.
The order of the Commonwealth Court is affirmed.
Justice ZAPPALA files a concurring opinion.