OPINION
{1} The State appeals from an order excluding evidence of Defendant’s blood alcohol content (BAC) tested four hours after the accident caused by Defendant. In State v. Silago, 2005-NMCA-100, ¶¶ 12-15, 23, 138 *309N.M. 301, 119 P.3d 181, 2005 WL 1994285 [No. 24,854 (N.M. Ct.App. June 27, 2005) ], filed today, we held that a trial court erroneously excluded evidence of the defendant’s BAC based on a flawed reading of our case law on the relation back of BAC evidence, and we remanded that case so the trial court could consider whether to admit or exclude the evidence under proper principles of law. This case provides us with an opportunity to review another trial court’s decision to exclude such evidence where the trial court appeared to use proper legal principles. Because the trial court did not abuse its discretion in performing its gatekeeping function, we affirm.
BACKGROUND
{2} The parties do not dispute that Defendant admitted to drinking two beers at 8:30 p.m., that the accident occurred at 11:30 p.m., and that a blood test four hours after the accident measured Defendant’s BAC at 0.10. Defendant moved in limine for exclusion of the BAC, arguing that the BAC four hours after the accident was relevant only if it tended to show what Defendant’s BAC was at the time Defendant was driving. Relying on State v. Baldwin, 2001-NMCA-063, 130 N.M. 705, 30 P.3d 394, Defendant argued that the State had to introduce some evidence providing a nexus between the BAC at the time of driving and the BAC four hours later. Id. ¶ 8. There was no behavior evidence suggesting that Defendant’s BAC was over the legal limit at the time of the accident, and Defendant argued that even scientific evidence purporting to extrapolate back from the BAC of 0.10 was unreliable.
{3} At the hearing on Defendant’s motion, Defendant presented the testimony of Dr. Edward Reyes, whom the trial court accepted as an expert in pharmacokinetics. Reyes testified that he could not determine Defendant’s BAC at the time of driving based on her BAC four hours later because there are many variables that would affect the calculation, such as adrenaline in Defendant’s system or food in her stomach, either of which would slow the absorption of alcohol, and Defendant’s metabolic rate. He explained that the body is absorbing alcohol when alcohol is in the stomach, and that in this absorption phase, the BAC is increasing. Once the stomach empties, the contents go through the small intestine, which is where about 90% of alcohol absorption takes place. When absorption is complete, the BAC peaks and then begins to decrease as the alcohol is metabolized. A surge of adrenaline, such as might occur during a car accident, or food in the stomach could delay absorption for up to four hours. Reyes concluded that uncertainty about these various factors and about Defendant’s metabolic rate prevented him from making even an educated guess about Defendant’s BAC at the time of driving.
{4} The State’s expert, Curtis Caylor, accepted by the trial court as an expert in toxicology, also testified at the hearing. Caylor stated that in order to determine Defendant’s BAC at the time of driving based on her BAC four hours later, a calculation known as retrograde extrapolation, one would have to make several assumptions. One would have to assume that Defendant had been drinking over a period of time, that she had reached her peak or plateau level, and that she had no alcohol to drink after the accident. Caylor agreed with Reyes that adrenaline or food in the stomach would slow alcohol absorption, but it would not stop it. Although Caylor could not provide a range of average absorption rates, he testified that average absorption time is from fifteen minutes to one hour after ingestion, regardless of the amount of alcohol in the stomach. Caylor did not give an opinion as to Defendant’s likely BAC at the time of driving, but a reasonable inference from his testimony was that it was higher than the BAC of 0.10 measured four hours after the accident.
{5} After hearing the testimony of Reyes and Caylor, the trial court stated that it found Reyes’s testimony “compelling” and ruled that it would not allow testimony on retrograde extrapolation at trial. The court entered a written order excluding the results of Defendant’s BAC test.
DISCUSSION
{6} The State argues that the trial court erroneously excluded Defendant’s BAC test results because it misapplied Baldwin. While the admission and exclusion of evidence is ordinarily within the sound discre*310tion of the trial court, “the threshold question of whether the trial court applied the correct evidentiary rule or standard is subject to de novo review on appeal.” State v. Torres, 1999-NMSC-010, ¶ 28, 127 N.M. 20, 976 P.2d 20.
{7} Baldwin and subsequent cases involve the question of what constitutes sufficient evidence of BAC at the time of driving when there is a delay between the time of driving and the testing of BAC.2001-NMCA-063, ¶ 2; State v. Christmas, 2002-NMCA-020, ¶ 19, 131 N.M. 591, 40 P.3d 1035; State v. Martinez, 2002-NMCA-043, ¶ 1, 132 N.M. 101, 45 P.3d 41. The thrust of the State’s argument is that this line of cases addresses only the sufficiency of the evidence and says nothing about the admissibility of BAC evidence, which is the issue in this case. We believe these cases provide context for determining the admissibility of BAC evidence.
■ {8} In any case where the State attempts to prove a violation of the per se DWI statute, which requires a minimum specific BAC at the time “[t]he defendant operated a motor vehicle,” UJI 14-4503 NMRA, the critical inquiry is how to determine the defendant’s BAC at the time of driving if there is a significant delay between the time of driving and the time BAC is measured. “[T]he longer the delay between the time of [the] incident and [the] sample collection, the more difficult it becomes, scientifically, to draw reasonable inferences from one data point, back to the driving time.” Christmas, 2002-NMCA-020, ¶ 20, 131 N.M. 591, 40 P.3d 1035 (internal quotation marks and citations omitted). Consequently, when the delay between driving and testing is significant, the state must prove a nexus between the defendant’s BAC score and the time of driving through evidence corroborating the inference that the defendant’s BAC at the time of driving was at the statutory level of 0.08 or above. See Baldwin, 2001-NMCA-063, ¶¶ 4, 14, 24, 130 N.M. 705, 30 P.3d 394 (reversing conviction where corroborative evidence was insufficient to support relation-back inference when the delay was two hours and fifteen minutes); Christmas, 2002-NMCA-020, ¶ 6, 131 N.M. 591, 40 P.3d 1035 (assessing corroborative evidence where the delay was “about an hour”); Martinez, 2002-NMCA-043, ¶¶ 11, 13, 132 N.M. 101, 45 P.3d 41 (evaluating corroborative evidence where the delay was one and a half hours). We have declined to provide an exhaustive list of the types of corroborative evidence that would suffice, but we have said that
[t]he evidence might include a police officer’s observation of significant incriminating behavior on the part of the driver, or the evidence might include expert testimony relating the test result back in time to the time of driving.
Baldwin, 2001-NMCA-063, ¶ 2, 130 N.M. 705, 30 P.3d 394. Thus, we have suggested that retrograde extrapolation might suffice to provide a nexus between the results of a defendant’s BAC test and the likely BAC at the time of driving.
{9} While our prior cases have not specifically addressed the admissibility of retrograde extrapolation evidence, those cases suggest that such evidence may well be admissible if it satisfies the usual requirements for admissibility of expert testimony. Here, however, the question is not squarely one involving the admissibility of expert testimony on retrograde extrapolation. Instead, the issue is whether Defendant’s BAC was admissible. But the admissibility of the BAC result is intertwined with the expert testimony on retrograde extrapolation because, under the Baldwin line of cases, the BAC taken four hours after the accident is relevant only if it tells the fact finder something about Defendant’s BAC at the time of the accident. Id. ¶ 15 (explaining that a jury cannot determine how a BAC result relates back to a particular BAC at the time of driving “[wjithout evidentiary guidance”). Thus, if an expert can testify as to a method that reliably extrapolates from a defendant’s BAC test result to a likely BAC at the time of driving, the BAC result is helpful to the fact finder and may be admissible.
{10} This brings us to a consideration of State v. Alberico, 116 N.M. 156, 861 P.2d 192 (1993). In that ease, our Supreme Court explained that “Rule 11-702 [NMRA] establishes three prerequisites for admission of expert testimony: (1) the expert must be qualified, (2) the scientific evidence must as*311sist the trier of fact, and (3) the expert may-only testify to scientific, technical or other specialized knowledge.” Lopez v. Reddy, 2005-NMCA-054, ¶ 11, 137 N.M. 554, 113 P.3d 377 [Vol. 44, No. 22 SBB 27 (June 6, 2005)] (citing Alberico, 116 N.M. at 166, 861 P.2d at 202 (internal quotation marks omitted)). Here, the parties do not dispute the first and third criteria. The trial court accepted both experts as qualified, and the defense expert did not appear to dispute the scientific validity of retrograde extrapolation, providing all variables are known. Consequently, the issue was the second criterion, whether retrograde extrapolation in the context of the specific facts of this case would be helpful to the fact finder. Alberico, 116 N.M. at 170, 861 P.2d at 206 (explaining that the second criterion requires consideration of the testimony’s “probative value, that is, whether it reliably and accurately proves what it purports to prove”).
{11} The defense expert testified that it was impossible to extrapolate back from Defendant’s BAC test result to a likely BAC at the time of driving because there were too many unknown variables, such as whether Defendant had food in her stomach and what Defendant’s metabolic rate was. The State’s expert testified that it was possible to estimate Defendant’s rate of alcohol burnoff, and this testimony could arguably give rise to the inference that Defendant’s BAC at the time of driving was higher than the 0.10 BAC four hours later. Thus, the trial court was faced with the conflicting testimony of two experts. While it is ordinarily for the jury to resolve conflicting testimony, here, the testimony of the State’s expert was so vague and general as to provide no real assistance to the trier of fact. Thus, under the second Alberico criterion, the trial court could properly conclude that the State did not meet its burden.
{12} Viewing the BAC result against the backdrop of the defense expert’s testimony, we cannot say that the trial court’s exclusion of the BAC was “obviously erroneous, arbitrary, or unwarranted.” Id. If, as the expert testified, retrograde extrapolation would be impossible under the circumstances, the BAC would have no meaning to the jury, especially given the absence of evidence that Defendant was behaving as if she was intoxicated at the time of the accident. There was simply no evidence providing a nexus between the BAC and the time of driving, and the BAC was therefore just a meaningless number. See Baldwin, 2001-NMCA-063, ¶ 15, 130 N.M. 705, 30 P.3d 394 (“Without evidentiary guidance, no jury of lay persons can know, beyond a reasonable doubt, whether a 0.08 BAC test result would equate to a BAC reading at the time of driving of 0.08, 0.12, 0.04, or anything in between.”). Thus, under the unique circumstances of this case, the trial court did not abuse its discretion in excluding the BAC as evidence of Defendant’s BAC at the time of driving.
{13} By this opinion, we do not mean to decide, and we are expressly not deciding, that retrograde extrapolation evidence is always inadmissible. See generally Mata v. State, 46 S.W.3d 902, 916 (Tex.Crim.App. 2001) (en banc) (stating, after exhaustive discussion of evidence in the case and scientific literature, court believed reliability of retrograde extrapolation could be proved in a given case, but expert must demonstrate understanding of difficulties and subtleties and must be able to explain testimony with clarity to trial court). Different experts may testify differently in future cases, or the same experts may testify differently in future eases. We do not mean to imply that our decision affirming the trial court’s exercise of discretion is one of law. Trial courts must review the evidence presented in each case anew and decide whether the tendered evidence will meet the Alberico test. In turn, appellate courts must review the exercise of discretion in each ease to determine whether it is erroneous, arbitrary, unwarranted, or an abuse.
CONCLUSION
{14} For the foregoing reasons, we affirm the trial court’s exclusion of Defendant’s BAC.
{15} IT IS SO ORDERED.
WE CONCUR: LYNN PICKARD and IRA ROBINSON, Judges.