SIMONETT, Justice.
This case involves the decision of a city council refusing to rezone certain land from single-family residential to multiple-unit residential and commercial and the order of the district court reversing that decision. The city appeals. We conclude the review procedure used by the district court was incorrect and we reverse and remand for trial.
The tract involved is about 15 acres of vacant, wooded land lying in the City of Coon Rapids. It is somewhat triangular in shape, bounded by Foley Boulevard on the northwest and Coon Rapids Boulevard on the southwest, with the eastern boundary, the base of the triangle, bounded by a park and some single-family dwellings. The adjacent land on the opposite side of Foley Boulevard is devoted to commercial use, and on the opposite side of Coon Rapids Boulevard is an industrial park. The boulevards, which intersect at the west end of the tract, are busy, arterial streets. Thus the property, as the trial court put it, is “at the confluence of a mixture of commercial, residential and industrial districts.”
Plaintiff-respondents, Laron, Galen and Carol Honn, purchased the property in 1968 and shortly thereafter petitioned to rezone about 6 acres to multiple-unit residential. The city’s planning commission recommended approval but expressed concern that multiple dwellings should not be allowed until sewer service became available. In November 1968 the city council unanimously approved the planning commission’s recommendation for the rezoning but no more was done and the ordinance was never amended.
In 1975 respondents filed a new application to rezone 14 acres from single-family residential (R-2) to multiple-unit residential (M) and one-half acre to general commercial (GC). This would allow construction of multiple-unit apartments on the 14 acres and various businesses on the half-acre parcel. At a March planning commission hearing, the commission presented its evaluation of the proposal. The commission stated the Honn development would be a buffer between adjacent commercial and residential zones and would be consistent with broader neighborhood plans and with the city’s “goals and policies.”
The commission recommended approval of the rezoning application.
When the recommendation reached the city council, it ran into strong citizen opposition. At the first reading of the proposed rezoning ordinance, neighbors expressed fears that the development would destroy the residential character of the area, that increased drainage needs would require increased taxes, and that local recreational facilities would be overburdened. Possible alternatives were discussed at the second reading in June, but in the end the city council voted 4 to 1 to deny rezoning.
Respondents Honn sued the city in August 1978, bringing a declaratory judgment action, alleging among other things that the denial of rezoning “exceeds the zoning power” of the city.
The trial court, instead of conducting a trial, treated the case as one “in the nature of certiorari”; it instructed the parties to prepare an “agreed upon record” of the proceedings before the city council and, after reviewing that record, issued its findings of fact and conclusions of law. The trial court concluded that the city’s denial of rezoning was arbitrary, capricious and unreasonable. It therefore ordered the city to rezone the property as recommended by the planning commission. The city appeals.
We are presented with two issues, broadly stated as: (1) Procedurally, could the trial court, in fashioning a method of review, use the certiorari approach; and (2) on the merits, should the city council’s decision have been reversed?
I.
When the parties appeared in district court with their witnesses for trial, the court advised them there would be no trial as such. Rather, the parties were told to agree upon a record of what had occurred before the city council and present this for review. This was done. The city assembled the various documents and studies prepared by the planning commission, the city staff and respondents, which had been presented at the city council meeting, plus a transcript of the tape recording (somewhat garbled) of the hearings themselves, and the city council’s findings of fact (64 detailed findings) and conclusions of law, prepared with the help of counsel and adopted by the council at a special meeting. It was this record, filling a large three-ring loose-leaf binder, that the trial court reviewed and on which it based its decision.
Appellant city argues it had no “trial” on its declaratory judgment action, that there was no true record, no opportunity to examine or cross-examine witnesses or to submit further testimony. Although both parties had originally protested the certiorari procedure, respondents now claim, naturally enough, that all relevant information was before the court and thus there was no need for a trial.
In using this procedure the trial court pointed out its only proper function was to review the
reasonableness
of the decision to deny rezoning. “[I]t would be a disservice to the City Council,” said the trial court, “if the Court entertained any evidence [submitted by plaintiffs] that was not before the City Council at the time and, thus, held them [the Council] to be arbitrary, capricious or unreasonable in determining the matter on evidence adduced before me and not before the Council.” In other words, the trial court took the position the only evidence relevant on the reasonableness of the council’s decision was evidence that had been presented to the council at its hearing.
Since the parties came to court with no record of the proceedings below, indeed with no findings of fact but only the minutes of its public hearing, the trial court instructed the parties to submit an “agreed upon” record. The trial court took the position that its limited scope of review dictated “exclusion of any evidence which was not considered by the legislative body at the time the zoning decision was made. A contrary ruling would result in a trial
de novo,
a result which is logically incompatible with the scope of review required.” Our first inquiry, then, is whether the trial court adopted a proper review procedure.
II.
Minn.Stat. § 462.361, subd. 1 (1980), provides:
Any person aggrieved by an ordinance, rule, regulation, decision or order of a governing body or board of adjustments and appeals acting pursuant to sections 462.351 to 462.364 may have such ordinance, rule, regulation, decision or order reviewed by an appropriate remedy in the district court, subject to the provisions of this section.
1. What, then, is the “appropriate remedy” for review? The statute does not say; but litigants have used a variety of remedies for a variety of zoning cases. The remedies include mandamus, certiorari, injunction and the declaratory judgment action.
In none of our zoning cases, as the trial court observed, do we either approve or disapprove of a review by certiorari. Certiorari was used in
Olsen v. City of Hopkins,
288 Minn. 25, 178 N.W.2d 719 (1970), to test denial of a building permit, and it was used in
Ram Development Co. v. Shaw,
309 Minn. 139, 244 N.W.2d 110 (1976), to review the validity of a certain voting procedure in adopting a rezoning ordinance. Both cases, however, involved issues best characterized as quasi-judicial.
Certiorari is a venerable common-law right, recognized also in this state by statute. Minn.Stat. ch. 606 (1980). It is issued by a court to review the proceedings of a tribunal exercising judicial or quasi-judicial functions. 14 C.J.S.
Certiorari
§ 1 (1939). It is considered an extraordinary remedy to redress obvious defects of justice for which no ordinary remedy is available. At common law it issued only where there was a manifest injustice and not where there was any doubt as to the petitioner’s rights. Its purpose, as section 606.01 puts it, is “to correct any proceeding.” As this court has stated:
[C]ertiorari lies to review the quasi-judicial acts and proceedings of a municipal body to determine — through an inspection of the record — if the body had jurisdiction, kept within it, and to examine the evidence, not for the purpose of weighing it, but to ascertain whether it furnished any legal and substantial basis for the action taken.
Beck v. City Council of St. Paul,
235 Minn. 56, 58, 50 N.W.2d 81, 82 (1951). Minnesota law is also consistent with common law in stating that certiorari is not appropriate to review legislative acts.
Mahnerd v. Canfield,
297 Minn. 148, 211 N.W.2d 177 (1973). Our case law has also established that on certiorari a court should only reverse when, as a matter of law, the evidence did not provide a substantial basis for the decision.
Haaland v. Pomush,
263 Minn. 506, 117 N.W.2d 194 (1962);
State ex rel. Spurck v. Civil Service Board,
226 Minn. 240, 32 N.W.2d 574 (1948).
2. Since certiorari is properly used only to review quasi-judicial issues and because rezoning involves a legislative determination, certiorari is not a proper remedy for review here. The distinction between what is quasi-judicial and what is legislative is not always easy to make, but the argument that rezoning should be treated as quasi-judicial was made and lost in
State, by Rochester Association of Neighborhoods v. City of Rochester,
268 N.W.2d 885 (Minn.1978). In that case we reaffirmed our holding that in adopting or amending a zoning ordinance, a municipality acts in a legislative capacity; this is true regardless of the size of the tract involved.
See also Sun Oil Co. v. Village of New Hope,
300 Minn. 326, 220 N.W.2d 256 (1974);
Beck v. City of St. Paul,
304 Minn. 438, 231 N.W.2d 919 (1975).
Moreover, the traditional standard of review for certiorari is inappropriate here,
i.e.,
whether the evidence provides a substantial basis for the decision. Ordinarily, too, the usual remedy on certiorari is simply to set aside the action taken by the lower tribunal; here, instead, the trial court ordered the city council to rezone.
3. While the foregoing considerations may show the inappropriateness of certiora-ri to review a rezoning matter, a more basic question remains. If, as the trial court says, the only function of the district court is to decide the reasonableness of the city council’s decision on the record it had before it, then something like certiorari does seem proper, for certiorari brings up the record below for review.
First of all, it should be remembered that the standard of review for legislative zoning decisions is narrow. “As a legislative act, a zoning or rezoning classification must be upheld unless opponents
prove that the classification is unsupported by any rational basis related to promoting the public health, safety, morals, or general welfare.”
State, by Rochester Association of Neighborhoods v. City of Rochester,
268 N.W.2d 885, 888 (1978). In other words, the test is a “rational basis” test. A city council has broad discretion in legislative matters, and even if the city council’s decision is debatable, so long as there is a rational basis for what it does, the courts do not interfere.
See, e.g., Beck v. City of St. Paul,
304 Minn. 438, 448, 231 N.W.2d 919, 925 (1975), and
Sun Oil Co. v. Village of New Hope,
300 Minn. 326, 334, 220 N.W.2d 256, 261 (1974).
4. In determining if there is a rational basis for the city council’s decision, should the reviewing court consider only the evidence presented to the council or is there a trial
de novo,
where new evidence may be adduced? Our cases have not dealt specifically with this question. In
Sun Oil,
for example, we alluded to a “trial transcript” that was “permeated with testimony and stipulations by which Sun Oil attempts to show that its subject property would make a more suitable site for a service station than the present Shell station site.”
Id.,
300 Minn. at 334, 220 N.W.2d at 261. In some jurisdictions, like Michigan, the trial court holds a
de novo
hearing, but on the record made before the zoning body; the trial court is free to draw its own conclusions from the evidence presented below but may not receive new evidence.
Abrahamson v. Wendell,
76 Mich.App. 278, 256 N.W.2d 613 (1977). On the other hand, in Indiana, the trial court in reviewing the actions of a zoning board may receive additional evidence to supplement the record “so long as it confines such evidence to the subjects covered in the hearing before the zoning board.”
Metropolitan Board of Zoning Appeals v. Gateway Corp.,
256 Ind. 326, 337, 268 N.E.2d 736, 743 (1971).
Prior to 1979 this court reviewed zoning matters like other civil actions. But in that year, in our review of a denial of a special use permit to Northwestern College in Arden Hills, we announced a new rule. We held the scope of review adopted in
Reserve Mining Co. v. Herbst,
256 N.W.2d 808, 822 (Minn.1977), for review of state agency decisions was appropriate in reviewing the decisions of local governing bodies in zoning matters.
Northwestern College
v.
City of Arden Hills,
281 N.W.2d 865, 868 (1979). We quoted from
Reserve Mining
that our function was “to make an independent examination of an administrative agency’s record and decision” and to arrive at our own conclusions “without according any special deference to the same review conducted by the trial court.”
Id.,
281 N.W.2d at 868.
Recently, in
C. R. Investments v. Village of Shoreview,
304 N.W.2d 320, 325 (Minn.1981), we applied the approach adopted in
Northwestern College
and directly reviewed the village’s decision in denying a special use permit to see if it was supported by the record.
While in
Northwestern College
we were announcing the kind of review this court would take on appeal from the district court’s ruling, our quotation from
Reserve Mining
suggests that the district court’s review would be of the record made before the local zoning body, the same as for a governmental agency. The difficulty with this, however, is that city councils and zoning boards do not ordinarily make records of their proceedings as complete and as formal as those of a state administrative agency or commission. Nor should we expect this to be so. Often the city council conducts its hearing in informal fashion, with no accurate verbatim record kept and with relatively few remarks in its minutes to suffice as its findings of fact and conclu
sions. Here the trial court sought to remedy the situation by requiring an after-the-fact record to be completed by the parties. In this particular case, this procedure did not prejudice the landowners, since the trial court found in their favor, but such an agreed upon record runs the risk inherent in any opportunity to rationalize or justify what one has done before. In
Zylka, supra,
we commented that “the failure of the council to record any legally sufficient basis for its determination at the time it acted made a prima facie showing of arbitrariness inevitable.”
Id.,
283 Minn, at 198, 167 N.W.2d at 50.
5. We conclude, therefore, that the proper procedure for review should be as follows:
1) Review of a decision on rezoning may be obtained by a declaratory judgment action.
Injunctive relief may also be requested. Certiorari is not appropriate.
2) In the declaratory judgment action, the parties are entitled to a trial.
3) Since the issue is whether there was a rational basis for the municipal body’s legislative decision, evidence presented to the city council is, of course, relevant. Ordinarily this evidence will include documents received by the city council, such things as maps, plans, surveys, studies and reports prepared by both the city staff and by the landowners. At the least, a summary of statements of interested persons at the
° hearing should be made at the time of the hearing by the city council and kept in the official file. This kind of evidence should ordinarily be admitted at trial by stipulation.
4) New or additional evidence may be received at the trial, but it must be relevant to the issues that were raised and considered before the municipal body. Witnesses may testify and be cross-examined, but their testimony must be relevant to the issues that were raised and considered below.
5)The municipal body need not necessarily prepare formal findings of fact,'but it must, at a minimum, have the reasons for its decision recorded or reduced to writing and in more than just a conclusory fashion. By failing to do so, it runs the risk of not having its decision sustained.
This procedure should be followed in presenting any zoning matter, whether legislative or quasi-judicial, for review to the district court and for subsequent review to this court. This is not to say that the form of the action by which the procedure is initiated need always be a declaratory judgment action. Mandamus has its place, and there may be a quasi-judicial proceeding presenting a legal question to which certiorari still lends itself.
See Ram Development Co. v. Shaw,
309 Minn. 139, 244 N.W.2d 110 (1976) (where the narrow issue was whether an amendment of the zoning ordinance requiring a two-thirds vote of all the members of the council was accomplished when there were four ayes, two abstentions and one pass).
III.
Before proceeding to the last issue before us, it will be helpful to consider briefly the standard of review.
1. Our case law distinguishes between zoning matters which are legislative in nature (rezoning) and those which are quasi-judicial (variances and special use permits). Even so, the standard of review is
the same for all zoning matters, namely, whether the zoning authority’s action was reasonable. Our cases express this standard in various ways: Is there a “reasonable basis” for the decision? or is the decision “unreasonable, arbitrary or capricious”? or is the decision “reasonably debatable”?
Nevertheless, while the reasonableness standard is the same for all zoning matters, the nature of the matter under review has a bearing on what is reasonable. In enacting a zoning ordinance or in amending an ordinance to rezone, the approach is legislative; what is involved is a kind of municipal planning in which a wide range of value judgments is considered. On the other hand, in granting or denying a special use permit, the inquiry is more judicial in character since the zoning authority is applying specific use standards set by the zoning ordinance to a particular individual use.
State, by Rochester Association of Neighborhoods v. City of Rochester,
268 N.W.2d 885, 889 (Minn.1978).
2. For rezoning the standard is whether the classification is reasonably related to the
promotion
of the public health, safety, morals or general welfare.
State, by Rochester Association of Neighborhoods,
268 N.W.2d at 888. This comes from Minn. Stat. § 462.357, subd. 1 (1980), which says, “For the purpose of promoting the public health, safety, morals and general welfare” a municipality may regulate land use and improvements thereto by zoning. But the approach is different in a special use permit case, where reasonableness is measured by the standard set out in the particular local ordinance, not the statute. For example, the special use standard in
C. R. Investments, Inc. v. Village of Shoreview,
304 N.W.2d 320, 323 (1981), is typical: “If the Village Council shall determine the proposed use will not be detrimental to the health, safety, morals, on [sic] general welfare of the community * * * the Village Council may grant a special use permit.” As we pointed out in
Zylka v. City of Crystal,
283 Minn. 192, 196, 167 N.W.2d 45, 49 (1969), an arbitrary denial of a special use permit may be found when “the requested use is compatible with the basic use authorized within the particular zone and does not endanger the public health or safety or the general welfare of the area affected or the community as a whole.”
3. In other words, in legislative zoning, the municipal body is formulating public policy, so the inquiry focuses on whether the proposed use promotes the public welfare. In quasi-judicial zoning, public policy has already been established and the inquiry focuses on whether the proposed use is contrary to the general welfare as already established in the zoning ordinance.
Consequently, the reviewing courts, in determining what is reasonable, should keep in mind that the zoning authority is less circumscribed by judicial oversight when it considers zoning or rezoning than when it considers a special use permit or a variance.
IV.
Respondents urge that, even if certiorari is not appropriate (and we hold it is not), there should be an affirmance of the lower court’s decision. They argue that all the facts necessary for the trial court to make its decision were before the court and that the trial judge, in effect, applied the proper “rational basis” standard of review. Consequently, respondents contend this court should look through form to substance and in the interests of judicial economy affirm. Respondents point out that the city was given an opportunity to bolster its case by compiling the “agreed upon” record and drafting its findings of fact long after the hearing. We hold the parties were entitled to make their record at a trial and we remand to the lower court for that purpose.
It is evident that the trial court carefully reviewed the record. It made detailed findings, and the court recognized in its memo-randa that if the city council’s decision was “fairly debatable,” it was to be affirmed. Nevertheless the court concluded the city council’s decision could not stand.
The city council made findings that, if rezoning were granted, there would be: (1) a substantial increase in population density; (2) a substantial increase in traffic; (3) increased safety hazards, particularly for children; (4) an adverse effect on already limited recreational facilities; (5) an adverse effect on the physical environment, especially on open space; (6) a detrimental effect on the existing single-family residential neighborhood; (7) a failure to give due recognition to the reliance of adjacent homeowners on the R — 2 zoning remaining unchanged; (8) an adverse effect on the value of surrounding homes; (9) an adverse effect on consistent land planning policies and practices; and (10) a failure to allow citizens a meaningful opportunity to participate in community decisionmaking.
Respondents, in turn, argue that the city posited the worst possible result, magnifying the size and thus the impact of respondents’ development. While the development would result in greater population density and increased traffic and safety problems, respondents argue that these consequences would, as a practical matter, be essentially the same for a well-designed apartment complex as for full development under the current zoning ordinance. Respondents say they would provide recreational facilities in their development to meet the increased need and that there would be adequate provision made for open space. As for the city’s argument that apartments would harm the character of the neighborhood, respondents counter that the neighborhood is already mixed, including a commercial strip and an industrial park, and that, indeed, the proposed apartment complex would provide a desirable buffer for the existing single-family dwellings. Respondents assert that lowering of property values was simply not proven nor was it shown that resident expectation in stability was impaired.
1. Several observations are pertinent. The city, in calculating the adverse effect of the proposed development, assumed a maximum development of up to 399 apartment units. The trial court found that plaintiffs’ feasibility study projected only a maximum of 180 to 200 units, and that, in any event, the city would still be able to impose restrictions on any building proposal to assure appropriate safety, convenience and amenities. An underlying problem was the question of who would pay for the sewer and water. If a sewer system were brought in, it would involve more than the 14 acres at issue, with the neighbors likely bearing some of the expense. It is also apparent that the neighbors preferred the land stay as it was — vacant, wooded, open space. Respondents argued that the concern about safety, open space and recreational facilities would also be present with the building of single-family dwellings under current zoning, but they seemed to suggest that unless apartments were built, financial constraints would prevent anything else being built. The trial court was of the view that, notwithstanding the city’s recital of legally sufficient reasons for its denial, a factual basis for those reasons was lacking, and that “the City’s findings and conclusions appear to be after-the-fact rationalizations for a decision intended primarily to placate local residents.”
This brief overview of the record and the contentions of the parties suggests some of the wide range of factors involved in making a legislative decision here as to whether or not to rezone. The standard of review, as we have said, is whether the decision of the Coon Rapids City Council not to rezone is reasonably related to the promotion of the public health, safety, morals and general welfare of the community.
2. The precise issue before us, however, because of the unusual posture of this case, is whether the “agreed upon” record permitted the parties to present their respective positions fully. We think not. To reverse the city council’s action
denies the appellant city its day in court. To affirm the city council’s action deprives respondents Honn of the same right. We conclude, therefore, the case should be remanded to the district court for trial.
3. Testimony at trial, for example, might bear on whether the city council’s decision was simply a response to neighborhood opposition, as the trial court found, or if it had secure evidentiary basis. The trial court recognized the city council was not required to follow the recommendation of the planning commission but concluded “when the Council fails even to address the planning considerations put forth by its own experts, the possibility of an unreasonable decision is necessarily raised.” Here again, testimony at trial may be enlightening. Also, .the original classification of this property as single-family residential is presumed to be well planned and intended to be more or less permanent.
See Sun Oil Co. v. Village of New Hope,
300 Minn. 326, 335, 220 N.W.2d 256, 261 (1974), quoting
Hardesty v. Zoning Board,
211 Md. 172, 177, 126 A.2d 621, 623 (1956). The burden is on respondents to show either some mistake in the original zoning or that the character of the neighborhood has changed to such an extent no reasonable use can be made of the property in its current zoning classification.
Sun Oil Co.,
300 Minn, at 337, 220 N.W.2d at 261-262. The trial court found, “The record as a whole establishes that single family housing is inappropriate for this parcel.” Here again, testimony directed at this issue may be helpful.
In other words, the parties should have the opportunity at a trial to explain or attempt to explain their actions, having in mind the evidence must be relevant to the issues considered before the city council. Only with this kind of record can the city council’s action be reviewed as contemplated by
Northwestern College v. City of Arden Hills,
281 N.W.2d 865 (Minn.1979).
Reversed and remanded for trial.
SHERAN, C. J., took no part in the consideration or decision of this case.