726 So.2d 762 (1999)
James SCOGGINS, Petitioner,
v.
STATE of Florida, Respondent.
No. 90,627.
Supreme Court of Florida.
January 21, 1999.
Richard L. Jorandby, Public Defender, and Christine Sciarrino and Peggy Natale, Assistant Public Defenders, Fifteenth Judicial Circuit, West Palm Beach, Florida, for Petitioner.
Robert A. Butterworth, Attorney General, and Celia Terenzio and Ettie Feistmann, Assistant Attorneys General, West Palm Beach, Florida, for Respondent.
ANSTEAD, J.
We have for review Scoggins v. State, 691 So.2d 1185 (Fla. 4th DCA 1997), which expressly and directly conflicts with the decision in Rodriguez v. State, 559 So.2d 678 (Fla. 3d DCA 1990). We have jurisdiction. Art. V, § 3(b)(3), Fla. Const. For the reasons expressed, we approve the opinion below.
PROCEEDINGS BELOW[1]
Following a jury trial, petitioner James Scoggins was convicted of possession of cocaine. During deliberations, however, the jury notified the trial court that it could not reach a verdict, stating, "We do not have a unanimous jury at this time and those who are in disagreement feel that they will not change their minds. What should we do?" Outside the presence of the jury, the trial judge asked each counsel if they objected to his questioning the jury as to its numerical division on the verdict. Neither counsel objected. The judge then addressed the jury, whereby the following colloquy occurred:
COURT: ... [D]o you think further deliberations would help at all?
[FOREPERSON]: There are those who feel that further deliberations would not help them.
COURT: Okay. Can I assume by that, that more than one personthe split is more than one person?
[FOREPERSON]: Yes.
COURT: So, in other words, at least four to two?
[FOREPERSON]: Yes.
COURT: Okay. And what about if I reset the deliberations until tomorrow, have you come back, you think that would serve any useful purpose?
[FOREPERSON]: You have to do what you feel is right.
COURT: Really, I don't wantthis is a very sensitive area, because I'm not allowed to make inquiry about a jury's deliberations, just not allowed to. So I can't ask you more than that.
If you as a foreperson are advising me that you think in any way that by resetting this until tomorrow, that could help this jury come to a decision, I will do it. If you think there's no wayif you want to talk to the other jurors, and if you think there's no way, then I'll declare a mistrial.
[FOREPERSON]: Am I allowed to express my personal feelings?
COURT: No.
[FOREPERSON]: Perhaps we should go back into the room, just decide whether or not we should meet tomorrow, and then come back out again.
After retiring to discuss the matter, the jury decided to return the next day, and, after further deliberations, it found Scoggins guilty of the possession charge.
On appeal, Scoggins argued that it was fundamental and reversible error for the judge to inquire into the jury's numerical division regardless of the lack of objection. The Fourth District, in an opinion by Judge Gross, disagreed, holding that although it was error for the judge to inquire into the jury's numerical division, such questioning did not rise to the level of fundamental error. Scoggins, 691 So.2d at 1188-89. In so holding, the district court expressed its disagreement with the holding in Rodriguez, which reversed the defendant's conviction under similar circumstances, relying upon the Supreme Court's decision in Brasfield v. United States, 272 U.S. 448, 47 S.Ct. 135, 71 L.Ed. 345 (1926). See Rodriguez, 559 So.2d at 679.
BRASFIELD
In Brasfield, the defendant was convicted in federal court for conspiracy to possess and transport intoxicating liquors, and the Supreme Court overturned the conviction because the trial judge inquired into the numerical split of the jury during its deliberations. Although the judge did not make reference to which number favored conviction, the Court concluded that such inquiry was error, requiring reversal:
We deem it essential to the fair and impartial conduct of the trial, that the inquiry itself should be regarded as ground for reversal. Such procedure serves no useful purpose that cannot be attained by questions not requiring the jury to reveal the nature or extent of its division. Its effect upon a divided jury will often depend upon circumstances which cannot properly be known to the trial judge or to the appellate courts and may vary widely in different situations, but in general its tendency is coercive. It can rarely be resorted to without bringing to bear in some degree, serious although not measurable, an improper influence upon the jury, from whose deliberations every consideration other than that of the evidence and the law as expounded in a proper charge, should be excluded. Such a practice, which is never useful and is generally harmful, is not to be sanctioned.
272 U.S. at 450, 47 S.Ct. 135.
The Supreme Court considered a similar, but separate, issue in Lowenfield v. Phelps, 484 U.S. 231, 108 S.Ct. 546, 98 L.Ed.2d 568 (1988), in resolving whether the trial court's combined acts of giving a supplemental charge and of polling the jury as to whether further deliberations would be helpful in reaching a verdict, unduly coerced the jury into finding the defendant guilty:
Although the decision in Brasfield was an exercise of this Court's supervisory powers, it is nonetheless instructive as to the potential dangers of jury polling.
Petitioner's attempt to fit the instant facts within the holding of Brasfield is, however, unavailing. Here the inquiry as to the numerical division of the jury was not as to how they stood on the merits of the verdict, but how they stood on the question of whether further deliberations might assist them in returning a verdict. There is no reason why those who may have been in the minority on the merits would necessarily conclude that further deliberation would not be helpful, or that those in the majority would necessarily conclude otherwise. The two questions are clearly independent of one another. We believe the type of question asked by the trial court in this case is exactly what the Court in Brasfield implicitly approved when it stated: "[An inquiry as to numerical division] serves no useful purpose that cannot be attained by questions not requiring the jury to reveal the nature or extent of its division."
. . . .
We hold that on these facts the combination of the polling of the jury and the supplemental instruction was not "coercive" in such a way as to deny petitioner any constitutional right. By so holding we do not mean to be understood as saying other combinations of supplemental charges and polling might not require a different conclusion. Any criminal defendant, and especially any capital defendant, being tried by a jury is entitled to the uncoerced verdict of that body. For the reasons stated we hold that there was no coercion here.
Id. at 239-41, 108 S.Ct. 546 (citations and footnotes omitted). Hence, Brasfield and Lowenfield resulted in different outcomes while treating related, but distinguishable, circumstances.
While Brasfield clearly condemned the practice of inquiring into the jury's numerical division and found such error to automatically mandate reversal without consideration of any other circumstances, the opinion was not grounded on any constitutional provision. Rather, as noted in Lowenfield, the decision was an exercise of the Supreme Court's supervisory powers over the federal courts. 484 U.S. at 240, 108 S.Ct. 546. As such, Brasfield's per se reversal approach is not binding on the states. See Cornell v. Iowa, 628 F.2d 1044, 1048 (8th Cir.1980) (holding that "under the principles of federalism," the better rule is to leave the decision of whether to allow inquiry into the jury's numerical division to the "supervising state court") (quoting Ellis v. Reed, 596 F.2d 1195, 1200 (4th Cir.1979); State v. Roberts, 131 Ariz. 513, 642 P.2d 858, 861 (1982); State v. Rickerson, 95 N.M. 666, 625 P.2d 1183, 1185 (1981)). Nevertheless, although not constrained by the rule in Brasfield, we still must determine whether we should adopt its holding for Florida courts. For the reasons well articulated in Judge Gross' opinion, we decline to adopt the per se approach of Brasfield.[2]
INQUIRY OF JURY
Consistent with Brasfield, the courts in this state which have addressed the issue hold that it is error for the trial judge to seek the jury's numerical division on the merits of the verdict when the jury has reached an impasse. See Rodriguez, 559 So.2d at 679; McKinney v. State, 640 So.2d 1183 (Fla. 2d DCA 1994).[3] In Rodriguez, the following colloquy occurred:
THE COURT: ... Ladies and gentlemen, you have been deliberating since about twenty to five. It's twenty after seven. This has been a three-witness case. Can I ask of you, please, can you tell me how far apart you are. I'm not asking whether you're voting guilty or not guilty, how far apart you are. In other words, five/one, four/two, three/three, that type of situation?
[THE FOREMAN]: At the present time we're probably four/two and it's not closed at this point.
. . . .
[A JUROR]: Your Honor, is there any way we can have the court reporter replay some of what we have heard?
THE COURT: I have already ruled on that.
559 So.2d at 679. Fifteen minutes after this colloquy occurred, the Rodriguez jury returned a verdict of guilty. The third district opinion quoted the rule in Brasfield and held that inquiry into the jury's numerical division was fundamental error requiring reversal. Rodriguez, 559 So.2d at 679, 680. However, the Rodriguez opinion also found error in the trial court's comment about the case being a "three witness" case, and in the refusal to honor the jury's request for a read back of testimony. Id. Similarly, the second district in McKinney found error in the trial court's polling the jury as to its numerical division on the verdict and in giving a modified and improper "dynamite" or Allen[4] charge. McKinney, 640 So.2d at 1187. Based on the combined effect of these two errors, the court held that the errors were not harmless. Id.
As noted in the district court's opinion here, other states are divided on whether judicial inquiry into a jury's numerical division constitutes error.[5] Some hold that inquiry into numerical division when the jury has reached an impasse is improper. See State v. Lautzenheiser, 180 Ariz. 7, 881 P.2d 339, 342 n. 5 (1994) ("[I]mproper influence by a judge upon a jury is always inappropriate and usually harmful."); Lowe v. People, 175 Colo. 491, 488 P.2d 559 (Colo.1971); People v. Eppinger, 293 Ill.App.3d 306, 227 Ill.Dec. 844, 688 N.E.2d 325, 329 (Ct.1997) (relying on People v. Golub, 333 Ill. 554, 165 N.E. 196 (1929)); State v. Hutchins, 43 N.J. 85, 202 A.2d 678, 684-85 (1964), overruled on other grounds by State v. Czachor, 82 N.J. 392, 413 A.2d 593 (1980); State v. Rickerson, 95 N.M. 666, 625 P.2d 1183, 1185 (1981); State v. Boogaard, 90 Wash.2d 733, 585 P.2d 789, 793-94 (1978); State v. McMahon, 186 Wis.2d 68, 519 N.W.2d 621, 630 (Ct.App.1994) (relying on Mead v. City of Richland Ctr., 237 Wis. 537, 297 N.W. 419, 422 (1941)); cf. White v. State, 95 Nev. 881, 603 P.2d 1063, 1065 (1979) (finding no error in judicial inquiry where no answer was given, conversation was limited to substantial changes in balloting during deliberations, and judge did not urge jury to reach a verdict).
Other courts hold that mere inquiry into the jury's numerical division, absent circumstances indicating coercion, does not constitute error. These courts reason that inquiry is permissible where it serves such purposes as determining the timing for recesses, whether there has been progress toward a verdict or agreement among the jurors, or whether the court should declare a mistrial when a jury is at an impasse. See Hardin v. State, 225 Ark. 602, 284 S.W.2d 111, 115 (Ark.1955); People v. Proctor, 4 Cal.4th 499, 15 Cal.Rptr.2d 340, 842 P.2d 1100, 1122 (1992), aff'd sub nom. Tuilaepa v. California, 512 U.S. 967, 114 S.Ct. 2630, 129 L.Ed.2d 750 (1994); Peppers v. State, 261 Ga. 338, 404 S.E.2d 788, 791 (1991); Sharplin v. State, 330 So.2d 591, 596 (Miss.1976); State v. Baker, 293 S.W.2d 900, 905 (Mo.1956). The State urges us to accept this approach.
However, as Judge Gross, writing for the majority in the opinion below, reasoned:
The fallacy in this approach is that it equates the state of numerical division with the stage of deliberations. For this reason we align ourselves with our sister courts, the federal courts and those state courts that have held that a trial judge should not inquire into the numerical division of the jury. For whatever reason, whether to gauge the time for an evening recess or to decide whether to give the jury a deadlock charge, if a trial judge inquires into the sensitive area of the possibility of a verdict, the better practice is to admonish the jury at the outset not to indicate how they stand as to conviction or acquittal.
The reasons for the rule precluding a judge from delving into the jury's numerical division are those articulated in Brasfield: the inquiry serves no useful purpose that cannot be attained through less intrusive questions; the inquiry has a tendency to be coercive; and it interferes with the proper relation of the judge to the jury. 272 U.S. at 450, 47 S.Ct. at 135-36. A principal aim of a jury trial is the receipt of a verdict that fairly reflects the considered judgment of each juror. See Fla. Std. Jury Instr. (Crim.) 2.09. Maintaining the secrecy in jury deliberations is important to insure an open and uninhibited exchange of ideas among the jurors. When combined with comments that belie the judge's feelings, or with instructions such as the jury deadlock charge, disclosure of the jury's numerical division risks conveying the message that the court believes that the majority should prevail, creating the "doubly coercive effect of melting the resistance of the minority and freezing the determination of the majority." [People v.] Wilson, [390 Mich. 689], 213 N.W.2d [193] at 195 [(Mich.1973)].
Scoggins, 691 So.2d at 1187-88 (citations and footnote omitted). We concur and agree with this reasoning.
We conclude that the potential for harm is inherent in an inquiry as to numerical division. The deliberations of a jury are extremely sensitive, and nothing should be done by the trial court to directly or indirectly influence a jury's deliberations, especially in a way that might divide or cast one juror or group of jurors against the others. Even ambiguous and subtle influences upon a minority quantified or identified by number can disturb the sensitive process by which jurors attempt to decide the case and resolve their differences. It is one thing for there to be a division that may occur and change naturally during the course of confidential deliberations. It is quite a different thing to identify and focus on the division in open court, outside the give and take exchange among jurors during their confidential proceedings. To permit otherwise would risk tainting the outcome of the trial by shifting the jury's focus from the law and evidence before it to the judge's motivation in inquiring into the status of a verdict. As Judge Gross noted, if the judge must inquire into the possibility of a verdict, "the better practice is to admonish the jury at the outset not to indicate how they stand as to conviction or acquittal." Scoggins, 691 So.2d at 1188.
FUNDAMENTAL ERROR
Even though we find error in the trial judge's inquiry, we also must determine whether such error amounts to fundamental error, thereby requiring reversal even in the absence of an objection at trial. As already noted, the court's inquiry here was conducted not only without objection, but with the apparent concurrence of defense counsel. Hence, we would have to determine that the error was of fundamental proportions to permit it to be asserted on appeal as a basis for ordering a new trial. For essentially the same reasons that we have rejected the per se rule of Brasfield, we agree with the district court that the error was not fundamental under the circumstances of this case.
The district court looked to the totality of the circumstances in determining whether the trial court's inquiry into numerical division required reversal:
If the totality of the circumstances supports the finding of improper coercion of the jury, then there has been a type of constitutional violation which is fundamental error, and per se reversible. On the other hand ... error not amounting to a constitutional violation is not fundamental error, so an objection at trial is necessary to preserve the issue and a harmless error analysis is appropriate.
Scoggins, 691 So.2d at 1189. The vast majority of state and federal courts reviewing state court proceedings have similarly rejected Brasfield's per se reversal approach in place of the "totality of the circumstances" test applied by the district court here.[6] Under this approach, convictions should be reversed only where the "cumulative effect of the trial court's actions had a coercive influence upon the jury." State v. Roberts, 131 Ariz. 513, 642 P.2d 858, 860 (1982). "What conduct amounts to coercion is particularly dependent upon the facts of each case." Id. Thus, the judge's inquiry must not be viewed in isolation, but rather, "in light of the context in which it was made." Locks v. Sumner, 703 F.2d 403, 407 (9th Cir.1983) (citing Ellis, 596 F.2d at 1200). We agree with this approach and the district court's application of it here in determining that the error was not fundamental.
This Court has defined fundamental error as "error that `reaches down into the validity of the trial itself to the extent that a verdict of guilty could not have been obtained without the assistance of the alleged error.'" Urbin v. State, 714 So.2d 411, 418 n. 8 (Fla. 1998) (quoting Kilgore v. State, 688 So.2d 895, 898 (Fla.1996)). We agree with the district court that Brasfield`s inflexible per se reversal approach should not be applied where, as here, the circumstances surrounding the inquiry do not indicate any actual improper influence on the jury. A finding of fundamental error is warranted only where the totality of the circumstance indicate a likelihood that the jury was actually improperly influenced by the court's inquiry into rendering a verdict. In that situation, the court's inappropriate inquiry would constitute fundamental error because its occurrence genuinely calls into question the basis upon which the jury's verdict was predicated. However, where the circumstances indicate no improper coercive influence on the jury's deliberations, such error should not be considered fundamental to the court's obligation to ensure a fair trial.
Here, the trial court asked the jury for its numerical split with the concurrence of Scoggins' counsel, and only after the court and counsel had been notified that the jury was unable to reach a unanimous verdict. The trial court then asked only if further deliberations would assist in reaching a verdict. It appears that the judge's inquiry related more to the issue of whether the jury should return the following day to continue deliberations than it did to the nature or extent of the jury's position on guilt or acquittal. The trial judge did not ask which jurors favored conviction, he acknowledged to the jury that he could not inquire into their numerical split, and he did not give supplemental instructions urging a verdict. The fact that the jury decided to return the next day to continue to deliberate also belies any suggestion that they were coerced into rendering a guilty verdict. See People v. Green, 91 Ill. App.3d 1085, 47 Ill.Dec. 590, 415 N.E.2d 595, 598 (Ct.1980).
As the district court observed, the inquiry in this case did not contain any suggestions of coercioni.e., the jury was not placed under time pressure to return a verdict; there was no exhortation of the jury to consider extraneous and improper factors, such as the government's fiscal health, in arriving at a decision; no potential holdout juror was isolated and demeaned for being in the minority; no charge indicated that the jury was required to reach a unanimous verdict or that the jurors had a duty to do so; there was no threat of marathon deliberations; and the judge did not ask whether the jurors in the majority favored acquittal or a guilty verdict, nor did he single out the minority jurors in imploring the jury to come to a decision. Scoggins, 691 So.2d at 1189. Rather, "[t]he judge's comments were balanced, encouraging neither acquittal nor conviction." Id. Hence, we conclude the error of the trial court in inquiring as to the numerical division of the jury did not constitute fundamental error.
CONCLUSION
In sum, we approve the district court's decision and opinion in full and hold that while it is error for trial courts to inquire into the jury's numerical division during deliberation, such error does not constitute fundamental error and require reversal where, as here, there is no objection, and the totality of the circumstances does not support a conclusion that there was a coercive influence on the jury's verdict. We disapprove Rodriguez to the extent it holds that such error is fundamental and always requires a new trial.[7]
It is so ordered.
HARDING, C.J., SHAW, WELLS and PARIENTE, JJ., and OVERTON and KOGAN, Senior Justices, concur.
NOTES
[1] This outline is derived from the opinion below. Scoggins v. State, 691 So.2d 1185 (Fla. 4th DCA 1997).
[2] Indeed, nearly every federal circuit court which has addressed the issue has rejected Brasfield `s per se reversal mandate when reviewing state court proceedings on federal habeas corpus petitions. See Lowenfield, 484 U.S. at 240 n. 3, 108 S.Ct. 546; see also Montoya v. Scott, 65 F.3d 405, 412 (5th Cir.1995); Jiminez v. Myers, 40 F.3d 976, 980 n. 3 (9th Cir.1993); Williams v. Parke, 741 F.2d 847, 851 (6th Cir.1984); Locks v. Sumner, 703 F.2d 403, 406 (9th Cir.1983); United States ex rel. Kirk v. Director, Dep't of Corrections, 678 F.2d 723, 727 (7th Cir.1982); Cornell v. Iowa, 628 F.2d 1044, 1047 (8th Cir.1980); Ellis v. Reed, 596 F.2d 1195, 1197, 1200 (4th Cir.1979).
[3] Compare Lewis v. State, 369 So.2d 667 (Fla. 2d DCA 1979) (reversing for new trial where trial court's modified Allen charge made reference to single juror who voted against conviction), with McElrath v. State, 516 So.2d 276 (Fla. 2d DCA 1987) (holding that judge's remarks to jury following Allen instruction, while not proper, were not so egregious as to amount to fundamental error).
[4] Allen v. United States, 164 U.S. 492, 17 S.Ct. 154, 41 L.Ed. 528 (1896).
[5] See also George R. Priest, Annotation, Propriety and Prejudicial Effect of Trial Court's Inquiry as to Numerical Division of Jury, 77 A.L.R.3d 769 (1977).
[6] See Lowenfield, 484 U.S. at 241, 108 S.Ct. 546; Montoya, 65 F.3d at 414; Jiminez, 40 F.3d at 980; Locks v. Sumner, 703 F.2d at 406; Cornell, 628 F.2d at 1048; Ellis, 596 F.2d at 1200; State v. Roberts, 131 Ariz. 513, 642 P.2d 858 (1982); People v. Green, 91 Ill.App.3d 1085, 47 Ill.Dec. 590, 415 N.E.2d 595 (Ct.1980); Pace v. State, 473 So.2d 167 (Miss.1985); Sharplin v. State, 330 So.2d 591 (Miss.1976); State v. Baker, 293 S.W.2d 900 (Mo.1956); White v. State, 95 Nev. 881, 603 P.2d 1063 (1979); State v. Rickerson, 95 N.M. 666, 625 P.2d 1183 (1981); State v. Bussey, 321 N.C. 92, 361 S.E.2d 564 (1987); State v. Boogaard, 90 Wash.2d 733, 585 P.2d 789 (1978); State v. Blessing, 175 W.Va. 132, 331 S.E.2d 863, 865 (1985). Although in the minority, several states continue to adhere to Brasfield`s per se reversal rule. See McCampbell v. Kentucky, 796 S.W.2d 596 (Ky.Ct.App.1990); People v. Wilson, 390 Mich. 689, 213 N.W.2d 193, 195 (1973) (holding inquiry constitutes per se reversal); Commonwealth v. Robinson, 102 Pa.Super. 46, 156 A. 582 (1931) (holding that inquiry constitutes reversible error); cf. Taylor v. State, 17 Md.App. 41, 299 A.2d 841 (Ct.Spec.1973) (citing Brasfield as controlling authority but considering facts and circumstances of case).
[7] As a second issue, Scoggins argues that the trial court erred in reinstructing the jury on the definition of "exclusive" in defining the possession of cocaine offense, thereby relieving the state of its burden to prove the essential elements of the charged offense. We decline to address Scoggins' second issue as it is beyond the scope of the conflict issue.