Dissenting opinion on motion for rehearing by:
PAUL W. GREEN, Justice,joined by Justices DUNCAN and ANGELINI.
This case is solely about whether a breath specimen showing a driver’s blood-alcohol concentration (BAC) in excess of 0.16 an hour after a traffic stop constitutes more than a scintilla of evidence that the driver’s BAC exceeded 0.10 an hour earlier. I would hold that it does not. The majority, however, departs from the controlling standard of review in concluding it was “reasonable” for the trial court to uphold an administrative driver’s license revocation on the record in this case. Accordingly, I respectfully dissent.
The pertinent statute is clear: the Texas Department of Public Safety (DPS) will administratively suspend the driver’s license of any adult if it determines “the person had an alcohol concentration of [0.10 or more] while operating a motor vehicle in a public place.” Tex. TRansp. Code Ann. § 524.012(b)(1) (Vernon 1999) (emphasis added). Thus, the vital fact to be determined is whether the person’s BAC exceeded 0.10 when he was driving; it is not whether the person was intoxicated, i.e. physically impaired, when he was driving. There is no direct evidence of this vital fact; therefore, we are concerned with whether it may reasonably be inferred that Mireles’ BAC exceeded 0.10 at the time he was stopped.
The evidence consists of the expert testimony of George A. McDougall, Jr., the Bexar County Breath Test Technical Supervisor, and the testimony of the arresting officer. McDougall testified regarding the reliability of the intoxilyzer, its proper operation in this case, the validity of the test results, and Mireles’ breath test results. No attempt was made to extrapolate the test results back to the time Mireles was operating his vehicle. The officer noted that at the time of the stop Mireles was impaired, exhibiting slurred speech, glassy eyes, and an odor of alcohol.
The controlling legal issues are not in dispute. It is agreed, for example, that in this review of an administrative determination we utilize the substantial evidence standard. See majority op. at 428. We further agree that substantial evidence review equates to no-evidence review. See id. (“Substantial evidence requires only more than a scintilla_”). And it is, of course, axiomatic that in order for there to be more than a scintilla of evidence, the evidence must be probative.
Furthermore, we agree there is no direct evidence bearing upon whether Mi-reles’ BAC exceeded 0.10 at the time he was driving, although we know he was impaired at that time. In this regard, we also agree that impairment can be consistent with conditions other than alcohol intoxication. See id. at 430-31 (“[S]ymptoms of alcohol intoxication can be caused by medical conditions other than alcohol intoxication....”). But, as noted, impairment is not the vital fact to be proved.
*432Our disagreement centers on whether the evidence that exists is probative of the vital fact to be proved; that is, does the intoxilyzer evidence and the evidence of impairment tend to prove that Míreles’ BAC exceeded 0.10 an hour before the breath specimen was obtained. The majority insists that the 0.16 intoxilyzer result was probative of the fact that Míreles’ BAC exceeded 0.10 at the time of the stop when considered along with evidence that Míreles was impaired. See id. at 429 (“We believe that alcohol concentration an hour or more after the stop is indeed probative of alcohol concentration at the time of the stop ... ”); id. at 430-31 (impairment, considered along with intoxilyzer evidence, was “a reasonable basis” to uphold suspension). I disagree, because in the absence of backward extrapolation evidence, or evidence that impairment equates to a BAC of 0.10 or higher, the facts do not support a reasonable inference of the vital fact.
We know, because it is scientifically established, that a particular BAC level one hour after a stop yields three possible facts regarding alcohol concentration at the time of the stop: that it was higher, lower, or the same. This is essentially admitted by the DPS. See Tex. Dep’t Pub. Safety, OperatoR Manual at 5-10 (“Unless all the variables [food, time, type of alcohol, size of each drink, type of mix, individual oxidation rate, and time of last drink] are known, the exact ethanol concentration at the time of the arrest cannot be accurately and precisely predicted. The alcohol concentration may be higher, lower, or the same.”). As a result, a one-hour old 0.16 intoxilyzer reading, without reliable extrapolation evidence, cannot be probative of a specific alcohol concentration an hour earlier. At best, it can mean only that at the time of the stop, Míreles’ BAC was higher than 0.16, lower, or the same. And there is, of course, no way to determine from this record just how much higher or lower it might have been. Moreover, there is nothing to show that one of the three possible facts is more probable than the other. Consequently, no one of them can reasonably be inferred. See $56,700 in U.S. Currency v. State, 730 S.W.2d 659, 662 (Tex.1987) (“When circumstances are consistent with either of two facts and nothing shows that one is more probable than the other, neither fact can be inferred.”).
Similarly, evidence of impairment, even if shown to be caused by alcohol intoxication, can mean one of three things: a BAC lower than 0.10, equal to 0.10, or higher than 0.10. It does not tend to prove that a BAC exceeded 0.10 at any particular point in time and thus is not probative of the vital fact to be proved.
In sum, a 0.16 BAC an hour after the stop amounts to no more than a scintilla of evidence to establish that Míreles’ BAC exceeded 0.10 at the time he was driving. This evidence is therefore legally insufficient to support the driver’s license suspension. The evidence of intoxication also fails to rise to more than a scintilla of evidence tending to prove that Míreles’ BAC equaled or exceeded 0.10 when he was stopped. And if the evidence is added together, as the majority has done in an attempt to bootstrap its conclusion, it remains legally insufficient — zero plus zero is still zero.
I dissent because I agree with Míreles that there was no evidence to support a finding that his BAC exceeded 0.10 at the time he was driving. The judgment below should be reversed and rendered.
The superseded en banc majority opinion written by Justice Sarah B. Duncan, previously issued on October 30, 1998, contains an excellent discussion of the relevant science and how it is treated in these types of cases by courts in this and other jurisdictions. Justice Duncan’s opinion is attached as an appendix and is adopted as a supplement to this opinion.
*433APPENDIX
No. 04-97-01007-CV
Albert MIRELES, Jr., Appellant v. TEXAS DEPARTMENT OF PUBLIC SAFETY, Appellee
From the County Court at Law No. 7, Bexar County, Texas
Trial Court No. 240,723
Honorable Timothy F. Johnson, Judge Presiding
Opinion by: Sarah B. Duncan, Justice (joined by Justices Green and Angelini)
Concurring opinion by: Phil Hardberger, Chief Justice (joined by Justice López)
Dissenting opinion by: Tom Rickhoff, Justice (joined by Justice Stone)
Sitting: Phil Hardberger, Chief Justice
Tom Rickhoff, Justice
Alma L. López, Justice
Catherine Stone, Justice
Paul W. Green, Justice
Sarah B. Duncan, Justice
Karen Angelini, Justice
Delivered and Filed: October 30,1998
REVERSED AND RENDERED
The issue presented in this case is exceedingly narrow but it will have tremendous ramifications for a vast number of cases: May an administrative law judge in an automatic license suspension proceeding find an alcohol concentration of 0.10 or more at the time of a stop from evidence establishing (1) at that time, the person was exceeding the speed limit and exhibiting slurred speech, glassy eyes, a strong odor of alcohol on his breath, and poor balance and (2) approximately one hour later, the person’s breath specimens indicated alcohol concentrations of 0.161 and 0.162?
Because of the importance of this issue, the members of this Court unanimously voted to consider this case en banc. We conclude the evidence, when considered in light of the reliable scientific data and relevant statutes, is legally insufficient to support an inference of an alcohol concentration of 0.10 at the time of the stop. Accordingly, we reverse the trial court’s judgment and render judgment reinstating Míreles’ driver’s license.
Factual and Procedural Background
Míreles was stopped at approximately 12:30 a.m. At that point in time, he was driving 76 m.p.h. in a 60 m.p.h. zone; he exhibited slurred speech, glassy eyes, a strong odor of alcohol on his breath, and poor balance; and he was unable to satisfactorily perform field sobriety tests. Mí-reles was therefore handcuffed and taken into custody. Approximately one hour later, Míreles provided two breath specimens. The first specimen, taken at 1:35 a.m., showed Míreles’ alcohol concentration to be 0.161. The second specimen, taken at 1:38 a.m., showed Míreles’ alcohol concentration to be 0.162. The Department therefore gave Míreles notice that it was suspending his license pursuant to chapter 524 of the Texas Transportation Code.
Míreles appealed. At the ensuing evi-dentiary hearing, over Míreles’ objection, the court admitted the arresting officer’s report and the affidavit and testimony of George A. McDougall, Jr., the Bexar County Breath Test Technical Supervisor', regarding the reliability of the intoxilyzer, its proper operation in this case, the validity of the test results, and Míreles’ breath test results. See Tex Transp. Code Ann. § 524.038 (Vernon Pamph.1997). At the conclusion of the hearing, Míreles argued “the Department has ... failed, as a matter of law, to prove its case” because “there is no evidence as to what the alcohol concentration of Mr. Míreles was at the time he was driving.” The ALJ disagreed and sustained the suspension, finding Mí-reles “was operating a motor vehicle in a *434public place, Loop 410, Bexar County, TX, with an alcohol concentration of 0.10 grams or greater of alcohol per 210 liters of breath ... as determined by Defendant’s submission to a breath test ... as requested.” Contending this finding was not supported by any evidence, Míreles appealed to the county court at law. The county court at law judge also disagreed and affirmed the suspension, finding that a 0.161 alcohol concentration at 1:35 a.m. constituted substantial evidence to support the ALJ’s finding of a 0.10 or more alcohol concentration at the time of the stop one hour earlier.
Míreles now appeals to this court, again contending there is no evidence to support the ALJ’s finding that he “was operating a motor vehicle in a public place, Loop 410, Bexar County, TX, with an alcohol concentration of 0.10 grams or greater of alcohol per 210 liters of breath ... as determined by Defendant’s submission to a breath test ... as requested.”
Administrative License Suspension
In Texas, if an adult is arrested for driving while intoxicated, and “submits to the taking of a specimen of breath or blood and an analysis of the specimen shows the person had an alcohol concentration of [0.10 or more],” the arresting officer must “serve or ... attempt to serve notice of driver’s license suspension by delivering the notice to the arrested person” and send a report to the Texas Department of Public Safety. Tex. Transp. Code Ann. § 524.011(b)(1) (Vernon Pamph.1997); Tex. Pen.Code Ann. § 49.01(2)(B) (Vernon 1994). The report must “(1) identify the arrested person; (2) state the arresting officer’s grounds for believing the person committed the offense; (3) give the analysis of the specimen if any; and (4) include a copy of the criminal complaint filed in the case, if any.” Tex. Tiiansp. Code Ann. § 524.011(c) (Vernon Pamph.1997).
Upon receipt of the arresting officer’s report, the Department “shall determine from the information in the report whether to suspend the person’s driver’s license.” Id. § 524.012(a) (emphasis added). If the person is an adult, the Department must suspend his license if it determines “the person had an alcohol concentration of [0.10 or more] while operating a motor vehicle in a public place.” Id. § 524.012(b)(1) (emphasis added); Tex. Pen.Code Ann. § 49.01(2)(B). Conversely, if “the person is an adult and the analysis of the person’s breath or blood specimen determined that the person had an alcohol concentration of a level below [0.10] at the time the specimen was taken,” “[t]he department may not suspend [the] person’s driver’s license.” Tex. Transp. Code Ann. § 524.012(c) (emphasis added). If the Department suspends the arrested person’s license, it must send a notice of suspension by certified mail. Id. § 524.013(a). However, the suspension does not take effect until the fortieth day after the date the arresting officer’s notice of suspension is received or presumed to have been received. Id. § 524.021(a).
If the Department makes the required finding and suspends the arrested person’s license, its determination is final unless the arrested person requests a hearing before an administrative law judge within fifteen days after receipt or presumed receipt of the notice of suspension. Id. §§ 524.012(d), 524.031. At the hearing, the Department must prove by a preponderance of the evidence, and the ALJ must affirmatively find, “the person had an alcohol concentration of [0.10 or more] while operating a motor vehicle in a public place.” Tex. Transp. Code Ann. § 524.035(a)(1)(A) (Vernon Pamph.1997) (emphasis added); Tex. Pen.Code Ann. § 49.01(2)(B).
Scope and Standard of Review in the Trial Court
If the ALJ makes the required findings and sustains the suspension, the arrested person may appeal the decision to a county *435court. Tex. Transp. Code Ann. § 524.041(a)-(b) (Vernon Pamph.1997). In this appeal, the scope of review is generally the agency record, and the standard of review is substantial evidence. Tex. Gov’t Code Ann. §§ 2001.174, 2001.175(e) (Vernon Pamph.1997); see also Tex. Transp. Code Ann. § 524.043(a)-(b) (Vernon Pamph.1997).
Substantial evidence review in the agency context is the functional equivalent of legal sufficiency review in the usual civil context; in both contexts, the issue is whether the challenged finding is supported by “more than a mere scintilla” of evidence. Compare Railroad Comm’n of Texas v. Torch Operating Co., 912 S.W.2d 790, 792-93 (Tex.1995) (“[s]ubstantial evidence requires only more than a mere scintilla .... ”), with Formosa Plastics Corp. USA v. Presidio Eng’rs and Contractors, Inc., 960 S.W.2d 41, 49 (Tex.1998) (legally sufficient evidence is more than a scintilla). Evidence is “more than a mere scintilla” if (1) there is direct evidence of the fact or (2) if there is no direct evidence of the “vital fact,” there is evidence from which the “vital fact” may reasonably be inferred. See Robert W. Calvert, “No Evidence” and “Insufficient Evidence” Points of Error, 38 Tex. L.Rev. 361, 363-65 (1960).
In this case, the Department concedes there is no direct evidence of the “vital fact,” i.e., Míreles’ alcohol concentration was 0.10 or more at the time he was stopped. Therefore, we are concerned with only the second aspect of quantitative legal sufficiency review — whether the “vital fact” may reasonably be inferred. This type of inquiry often presents “a very close question” and “it is not surprising to find' ... disagreement in a particular case.” Id. at 364. As a result, “[i]t is in this situation that the courts have needed and have tried to evolve a guiding rule of decision.” Id. This “guiding rule of decision” may be summarized as follows:
Evidence is legally insufficient “if reasonable minds cannot differ from the conclusion that [it] lacks probative force.” Id. Reasonable minds cannot differ that evidence lacks probative force if, “viewing the evidence in its most favorable light in support of the finding of the vital fact, considering only the evidence and the inferences which support the finding and rejecting the evidence and the inferences which are contrary to the finding,” 1 the evidence “is so weak as to do no more than create a mere surmise or suspicion of its existence.” Id. at 363-64. This standard is not met as a matter of law when the evidence is such that (a) it yields “equally reasonable and plausible” opposing inferences and (b) the “vital fact” cannot be inferred without “piling inference upon inference.” Id. at 364-65.
Standard of Review in the . Appellate Court
Substantial evidence review presents a question of law. See Firemen’s & Policemen’s Civil Serv. Comm’n v. Brinkmeyer, 662 S.W.2d 953, 956 (Tex.1984). We therefore review the trial court’s conclusion that substantial evidence supports the ALJ’s findings de novo. See In re Humphreys, 880 S.W.2d 402, 404 (Tex.), cert. denied, 513 U.S. 964, 115 S.Ct. 427, 130 L.Ed.2d 340 (1994). The interpretation of a statute is also a question of law and also, therefore, reviewed on appeal under a de novo standard. See Mitchell Energy Corp. v. Ashworth, 943 S.W.2d 436, 437 (Tex.1997).
Waiver
The Department first argues Míreles waived his points of error by failing to file *436an adequate brief. We disagree. Míreles’ brief substantially complies with the briefing rules, “acquaint[s] the court with the issues,’’and “presents] argument that will enable the court to decide the case.” Tex. R.App. P. 38.9. No more is required.
Substantial Evidence
The parties’ arguments are twofold. First, the parties dispute whether the evidence introduced at the hearing reasonably yields an inference of the “vital fact,” ie., Míreles’ alcohol concentration at the time of the stop was 0.10 or more. Second, the parties disagree upon whether the statutory framework for administrative license suspension creates a rebuttable presumption that the arrested person’s alcohol concentration at the time of the test is evidence of his alcohol concentration at the time of the stop. Although it might be more customary to address the statutory interpretation issues before addressing the evidentiary arguments, we address the arguments in the order presented. Understanding the evidentiary arguments and' the related science is dispositive of the evidentiary arguments and a necessary precursor to understanding the proper interpretation of the statutory framework for administrative license suspension.
Speeding
At the time Míreles was stopped, he was exceeding the maximum legal speed limit. However, people exceed the legal speed limit for a variety of reasons, most of which have nothing to do with intoxication. Therefore, to infer the “vital fact” — Mí-reles’ alcohol concentration was 0.10 or more at the time of the stop — from the evidence that he was, shortly before the stop, exceeding the legal speed limit would be patently unreasonable. As a result, evidence that Míreles was speeding cannot constitute the “more than a mere scintilla of evidence” required to support the ALJ’s finding. But cf. Martin v. Texas Dep’t of Pub. Safety, 964 S.W.2d 772, 775-76 (Tex. App.-Austin 1998, no pet. h.) (holding that reckless driving constitutes evidence of “prohibited alcohol concentration”).
Physical Signs of Intoxication
At the time Míreles was stopped, he exhibited slurred speech, glassy eyes, a strong odor of alcohol on his breath, and poor balance, all classic physical signs of intoxication, and he failed to perform the field sobriety tests satisfactorily. However, as the Department has recognized, a person may exhibit physical signs of intoxication for a reason or reasons completely unrelated to alcohol consumption. Texas Department of Public Safety, Texas Breath Alcohol Testing Program Operator Manual at 5-16 (TLE/br-38 (Rev.9/96)) (“Certain illnesses, diseases, or other drugs are able to produce symptoms similar to ethanol intoxication. Untreated diabetics, epileptics or trauma victims can all exhibit symptoms similar to ethanol intoxication.”) [hereinafter Tex. Dep’t Pub. Safety, Operator Manual]. And, as the Department also recognizes, these physical signs of intoxication may be exhibited by a person with an alcohol concentration of less than 0.10. Id. at 5-14 — 5-15.
Because physical impairment is consistent with conditions unrelated to alcohol and with alcohol concentrations above and below 0.10, it cannot reasonably yield an inference of the “vital fact” — Míreles’ alcohol concentration was 0.10 or more at the time of the stop. See McLean v. Moran, 963 F.2d 1306, 1307 (9th Cir.1992) (person can fail all field sobriety tests even if blood alcohol concentration is less than 0.10%); Commonwealth v. Loeper, 541 Pa. 393, 663 A.2d 669, 673 (1995) (“[I]mpairment evidence is not relevant since it does not logically or reasonably tend to prove or disprove that a defendant’s blood alcohol level was .10% or greater at the time that she drove her automobile, it does not tend to make such a fact more or less probable and it does not afford a basis for or support a reasonable inference or presumption regarding whether a defendant’s blood *437alcohol level was .10% or greater.”); Commonwealth v. Gonzalez, 519 Pa. 116, 546 A.2d 26, 83 (1988) (“It cannot reasonably be inferred from this evidence [of physical impairment] that ... [the appellant] was driving with any particular percentage of alcohol in his blood.”). But see Martin, 964 S.W.2d at 775-76 (holding that evidence of physical signs of intoxication and failed sobriety tests constitute “evidence suggesting ... the prohibited alcohol concentration”); Daricek v. State, 875 S.W.2d 770, 778 (Tex.App.-Austin 1994, pet. ref'd) (“[E]vidence of [the defendant’s] failure to pass field sobriety tests immediately after driving his vehicle tends to make it more probable that the failed blood or breath test taken an hour later accurately reflect [sic] the driver’s condition at the time of the offense.”).
Breath Test Results
Approximately one hour after the stop, Míreles submitted a breath specimen, which indicated his alcohol concentration at that point in time was 0.161. Three minutes later, Míreles submitted a second breath specimen, which indicated his alcohol concentration at that point in time was 0.162. No further breath, blood, or urine samples were taken, and the record is silent as to Míreles’ weight, what he drank, when and how fast he drank it, what he last ate and when, and whether he was a light, moderate, or heavy drinker. We are left then to decide whether Míreles’ alcohol concentration at the time of the stop can reasonably be inferred from evidence of his alcohol concentration one hour later. To decide this question requires at least an elementary understanding of the “pharma-cokinetics” of alcohol, that is its “absorption, distribution, biotransformation and excretion.” Kurt M. Dubowski, Absorption, Distribution and Elimination of Alcohol: Highway Safety Aspects, Supp. 10 J. Stud, on Alcohol 98, 98 (1985) [hereinafter Dubowski].2
So long as alcohol “remains in the stomach it does not affect the individual’s functioning and behavior. It is only when it is absorbed into the blood that it reaches the nervous system and produces its characteristic effects.” Richard F. Fitzgerald & David N. Hume, The Single Chemical Test for Intoxication: A Challenge to Admissibility, 66 Mass. L.Rev. 23, 28 (1981) [hereinafter Fitzgerald & Hume]. The difficulties associated with blood tests, however, “have led to the proliferation of breath-testing devices,” which can to some extent convert breath test results to blood test results; however, these tests are usually not administered until an hour or more *438after the stop. See id. at 23-24. Accordingly, “[i]n alcohol-related matters, such as driving under the influence, the most common toxicological question is: Given a blood alcohol concentration (BAC) at a specific time, can the BAC at an earlier time be predicted reliably?” Mark J. Reasor & Mark R. Montgomery, Driving Under the Influence: Is Retrograde Extrapolation of Blood Alcohol Scientifically Valid?, 9 W. Va. Law. 14, 14 (1996). The answer is “yes” — a process known as retrograde extrapolation can provide an answer if one knows certain information. See id. However, “[s]everal factors influence the value predicted by retrograde extrapolation, the largest uncertainty being the drinking pattern before the incident. The extent and reliability of the information available, and whether the incident occurred during the absorptive phase (where alcohol is still in the stomach) or postabsorptive phase (where all of the alcohol is in the blood) will determine the range of BACs that can be predicted.” Id. When this information is not available, retrograde extrapolation can only proceed if certain assumptions are made. Id.
In years past, one of the assumptions that “gained wide acceptance among police ‘technicians,’ police laboratory chemists, prosecutors, defense lawyers and trial judges” “is that alcohol is rapidly absorbed and that a peak BAC will be reached shortly after the last drink is taken.” Fitzgerald & Hume, 66 Mass. L.Rev. at 24. “A second assumption, which follows from the [first], is that if a suspect is arrested within the first half hour after he stopped drinking, he will be either at or very near his peak BAC at that time.” Id. at 25. “[A] third assumption is that once drinking has stopped a slow but steady decline in the BAC begins to take place immediately and will continue until all of the alcohol has been removed from the system.” Id. As a result, “[i]t is also assumed that a test within a ‘reasonable time’ after the incident is strong evidence as to the defendant’s state of intoxication at the time of the offense.” Id. The sum total of this chain of assumptions is the assumption that the arrested person’s alcohol concentration “at the earlier time of the incident was at least as high, and probably higher” as at the time of the test. Id. at 28; see also Jennifer L. Pariser, Note, In Vino Veritas: The Truth About Blood Alcohol Presumptions in State Drunk Driving Law, 64 N.Y.U. L.Rev. 141, 149-50 (1989) (“State legislatures and courts often ... presume that the defendant’s BAC at the time of the test will be lower than his BAC at the time of driving.”) [hereinafter Pariser]; Mullan v. State, 668 S.W.2d 427, 428 (Tex.App.-Texarkana 1984, no writ) (rejecting challenge to admissibility of breath test results in DUI proceeding without retrograde extrapolation because breath test administered “within forty-five minutes of the alleged offense,” “[t]he test result showed 0.13 per cent of alcohol by weight,” and expert testimony established “that alcohol absorption is complete within one-half hour to one hour”).
Today, however, “[m]ost experts agree that it ordinarily takes forty-five to ninety minutes to attain a peak BAC level on an empty stomach, and two to three hours if alcohol is consumed with or after a meal, while a few contend that the time lag between alcohol consumption and absorption into the blood stream is even longer.” McLean, 963 F.2d at 1309-10; see, e.g., Dubowski, Supp. 10 J. Stud, on Aloohol at 99. Accordingly, an alcohol concentration “measured some time after a driver’s arrest actually may be higher than if the test had been administered at the time of arrest, before the peak BAC had been reached.” McLean, 963 F.2d at 1310; see also id. at n. 2 (citing Loomis, Blood Alcohol in Automobile Drivers: Measurement and Interpretation for Medicolegal Purposes, 35 QuaRt. J. Stud. Alcohol, 458, 463-64 (1974)).
For instance, controlled experiments have demonstrated that peak alcohol concentration can occur within an hour of the *439rapid ingestion of straight 86-proof whiskey; but, if the same liquor is consumed slowly by a lighter person, the peak alcohol concentration may not occur for one hour or more. Dubowski, Supp. 10 J. Stud, on Alcohol at 104-105 (Fig. A & F). Similarly, while the rapid ingestion of beer may produce a peak alcohol concentration well within an hour, the rapid ingestion of champagne may not produce a peak alcohol concentration until several hours after the last drink. Id. (Fig. D & E). Perhaps most significantly in the context of this case, depending upon the type of alcohol consumed, the rate of consumption, and various other factors, a person’s alcohol concentration can increase over .06 of a gram less than forty-five minutes after a person stops ingesting alcohol, see id. (Fig.D), and it “can easily rise from a legal level to a criminal level in the time it takes to administer a test.” Pariser, 64 N.Y.U. L.Rev. at 151; see Fitzgerald & Hume, 66 Mass. L.Rev. at 32. In short, because so many factors are at work in the absorption and elimination of alcohol, and because so many of these factors are unknown, particularly in the DUI context, where there are constitutional protections against self-incrimination, “no forensically valid forward or backward extrapolation of blood or breath alcohol concentrations is ordinarily possible in a given subject and occasion solely on the basis of time and individual analysis results.” Dubowski, Supp. 10 J. Stud, on Alcohol at 106.
The reliable scientific data thus establishes that Mireles’ 0.16 breath test results one hour after the stop yield three possible inferences regarding his alcohol concentration at the time he was stopped: it was higher, lower, or the same. See, e.g., People v. Victory, 166 Misc.2d 549, 631 N.Y.S.2d 805, 811 n. 14 (N.Y.Crim.Ct.1995) (“[A] BAC level obtained an hour or two after operating a motor vehicle can be equally consistent with two contrary propositions. The test value could indicate a BAC of the same value or higher at the earlier time or it could indicate a much lower ‘and perhaps exculpatory BAC value at the time’ of driving.”) (citing and quoting Fitzgerald & Hume, 66 Mass. L.Rev. at 32); see also Tex. Dep’t Pub. Safety, OperatoR Manual at 5-10 (“Unless all the variables [food, time, type of alcohol, size of each drink, type of mix, individual oxidation rate, and time of last drink] are known, the exact ethanol concentration at the time of arrest cannot be accurately and precisely predicted. The alcohol concentration may be higher, lower, or the same.”). But see Martin, 964 S.W.2d at 775-76 (without reviewing reliable scientific evidence, court holds retrograde extrapolation not required and evidence of 0.19 breath test results taken approximately one and one-half hours after stop constitutes “evidence suggesting ... the prohibited alcohol concentration” at time of stop).
The Department insists the ALJ could calculate that Mireles’ alcohol concentration at the time of the stop was above 0.10 because the ALJ “knew,” “either through evidence or common knowledge,” “(1) one hour after driving Mireles had an alcohol concentration of .16; (2) he had not consumed alcohol between the time he was driving and the time of the breath test; and (3) during this time he was eliminating alcohol, that is, his alcohol concentration was declining.” We agree the ALJ “knew” the first and second facts because they are established by the evidence. But we cannot agree the ALJ “knew” the third factor because it is not established by the evidence, and it is not a matter of “common knowledge.” To the contrary, whether Mireles’ alcohol concentration was declining between the stop and the breath tests is the unknown fact that, if known, would be dispositive. As Professor Fitzgerald and Mr. Hume explained in their 1981 article:
A relatively high BAC value reflected in a single test (i.e., a 0.15%) would be strong evidence of the guilt of a defendant if it were known that both the offense and the sample occurred after *440the peak BAC had been obtained. The same BAC value, however, would be strong evidence of the innocence of the defendant if it were known that during the hour following arrest, or most of it, the BAC continued to rise from a low or exculpatory level to the value later reflected by test. Since either value (the higher prohibited value or the lower exculpatory value) would be equally consistent with the result of the later single test, its admission as presumptive proof of intoxication at the time of the offense is clearly unsupportable.
Fitzgerald & Hume, 66 Mass. L.Rev. at 32 (emphasis added).
“When circumstances are consistent with either of the two facts and nothing shows that one is more probable than the other, neither fact can be inferred.” Litton Indus. Prods., Inc. v. Gammage, 668 S.W.2d 319, 324 (Tex.1984). Therefore, because the rehable scientific data establishes Míreles’ alcohol concentration at the time of the stop might have been higher or lower than his subsequent breath test results and the record is silent as to any factor that might make one of these possible inferences more probable than the other, Míreles’ breath test results will not support a logical, rational, or reasonable inference of an alcohol concentration of 0.10 or more at the time of the stop. See Commonwealth v. Jarman, 529 Pa. 92, 601 A.2d 1229, 1231 (1992) (A blood test performed approximately one hour after stop and indicating a blood alcohol content of 0.114 “was no evidence upon which the expert could offer an opinion as to whether appellant’s blood alcohol level was in fact greater than or equal to 0.10% at the time of driving.”); Commonwealth v. Modaffare, 529 Pa. 101, 601 A.2d 1233, 1235-36 (1992) (A blood alcohol test performed one hour and fifty minutes after an accident and indicating a blood alcohol content of 0.108 was “no evidence that [appellant’s] blood alcohol level was equal to or above 0.10% at the time of the accident.”); Gonzalez, 546 A.2d at 34 (“ ‘Thus, if [an accused’s] drinking is confined to a period immediately before arrest, a test within the first 30 minutes after the arrest may show a low blood alcohol content whereas a test within [90] minutes after arrest may indicate [a much higher] blood alcohol level.’ ”) (quoting Commonwealth v. Speights, 353 Pa.Super. 258, 509 A.2d 1263, 1266 (1986), app. denied, 517 Pa. 594, 535 A.2d 83 (1987)).
To support its argument to the contrary, the Department relies upon Forte v. State, 707 S.W.2d 89 (Tex.Crim.App.1986), in which the Texas Court of Criminal Appeals held a jury may find an alcohol concentration of 0.10 or more “at the time of the offense” from chemical test result “near the time of the offense ” if it is convinced beyond a reasonable doubt “that the chemical test provides trustworthy evidence of alcohol concentration in a defendant’s breath, blood or urine” and “that an inference can be made from the results of the chemical test that the defendant had a 0.10% alcohol concentration in his body at the time of the offense.” Id. at 94-95; see also Dean G. Zioze, Comment, Trier of Fact May Infer Defendant’s Blood Alcohol Concentration at Time of Driving from Results of Subsequent Breathalyzer Test, 28 Suffolk U.L.Rev. 465, 467-68 (1994) (majority of jurisdictions, by statute or case law, permit a fact finder to infer a 0.10 or more alcohol concentration at the time of the stop from a breath test indicating an alcohol concentration in excess of 0.10 within one, two, three, or even four hours after the stop); id. at 468 n. 17 (citing cases); cf. Commonwealth v. Yarger, 538 Pa. 329, 648 A.2d 529, 531-32 (1994) (blood alcohol test result of 0.18% performed approximately forty minutes after stop sufficient for jury to infer BAC of 0.10% at time of stop, and state not required to present expert testimony to relate BAC test back to time of stop). But cf. Hartman v. State, 946 S.W.2d 60 (Tex. Crim.App.1997) (standard for admissibility of scientific expert testimony adopted in Kelly v. State, 824 S.W.2d 568 (Tex.Crim. *441App.1992), governs admissibility of retrograde extrapolation).
The Department’s reliance on Forte and similar cases is misplaced, however. As discussed above, the permissive inference recognized in Forte is only one of two possible inferences, and the reliable scientific data does not establish that one inference is more probable than the other, at least not when the record is devoid of evidence tending to establish the arrested person is in the elimination, rather than absorption, phase.3 As a result, the governing civil standard of legal sufficiency review prohibits the inference the Department requires to fulfill its burden to prove by a preponderance of the evidence that Míreles’ alcohol concentration was 0.10 or more at the time of the stop. Accordingly, whether the inference was permitted in criminal DUI cases at the time Forte was decided, and whether the inference will be permitted post-Hartman, are immaterial to the civil legal sufficiency analysis required in this case.
Because the record does not establish or even suggest Míreles was eliminating alcohol at the time of the stop, it is as plausible that his alcohol concentration was either higher or lower at the time of the stop. Which of these two plausible inferences is more probable cannot be established on this record viewed in light of the rehable scientific data. Accordingly, Míreles’ breath test results one hour after the stop are no evidence of an alcohol concentration of 0.10 or more one hour earlier. Nor can this vital fact be reasonably inferred from the evidence that Míreles was speeding and physically impaired; neither tends to establish a prohibited alcohol concentration. The record is therefore bare of evidence or reasonable inferences to support the ALJ’s finding that Míreles was driving in a public place with an alcohol concentration of 0.10 or more.
Statutory Framework
Anticipating this holding, the Department argues the evidentiary void is filled by the statutory framework, which implicitly permits the Department to suspend a license with proof of the subsequent breath test results alone. We disagree.
The objective of statutory interpretation and construction “is to determine and give effect to the Legislature’s intent.” Liberty Mut. Ins. Co. v. Garrison Contractors, Inc., 966 S.W.2d 482, 484 (Tex.1998). To accomplish this purpose, we look first “to the plain and common meaning of the statute’s words” when viewed in the context of the statute as a whole. Id. ‘When a statute is clear and unambiguous, courts need not resort to rules of construction or extrinsic aids to construe it, but should give the statute its common meaning.” St. Luke’s Episcopal Hosp. v. Agbor, 952 S.W.2d 503, 505 (Tex.1997). As the Supreme Court of Texas recently reiterated: “‘Courts must take statutes as they find them. More than that, they should be willing to take them as they find them. They should search out carefully the in-tendment of a statute, giving full effect to all of its terms. But they must find its intent in its language and not elsewhere .... They are not responsible for omissions' in legislation.’ ” Id. (quoting Simmons v. Arnim, 110 Tex. 309, 220 S.W. 66, 70 (1920)). Nonetheless, we are directed to “consider the object to attain, the circumstances of the statute’s enactment, legislative history, former statutory and common law, and the consequences of a particular construction.” Mitchell Energy Corp. v. Ashworth, 943 S.W.2d at 438. And “the construction of a statute by an agency charged with its execution is enti-*442tied to serious consideration unless the agency’s construction is clearly inconsistent with the Legislature’s intent.” Texas Water Comm’n v. Brushy Creek Mun. Util. Dist., 917 S.W.2d 19, 21 (Tex.1996).
Other state legislatures have enacted civil license suspension statutes and criminal DUI statutes expressly mandating or permitting a presumption that a person’s alcohol concentration at the time a specimen is taken is the same as or lower than his alcohol concentration at the time of the stop.4 This presumption is not, however, expressly mandated by the Texas administrative license suspension statutes, as the Department recognizes. Rather, the Department argues, several provisions in Texas’ administrative license suspension statutes implicitly indicate the legislature intended this rebuttable presumption to apply. We disagree.
In support of its argument, the Department points first to the statutory provisions requiring the Department to decide whether to suspend the arrested person’s license solely from the arresting officer’s report. See Tex. TRánsp. Code Ann. § 524.012(a)-(b) (Vernon Pamph.1997). Because the statute does not require this report to contain a retrograde extrapolation, see id. § 524.011(c), the Department argues it evidences a legislative intent to make the subsequent breath test results determinative of the arrested person’s alcohol concentration at the time of driving. This may or may not be true — but, if it is true, it is true only at the stage in which the Department is required to decide whether to send a notice of suspension. Nothing in the statutes indicates or even suggests the “administrative convenience” permitted at this preliminary stage was intended to relieve the Department of the burden expressly placed upon it by the Texas Legislature at the administrative hearing — to prove the arrested person’s alcohol concentration was 0.10 or more “while operating a motor vehicle in a public place.” Id. § 524.035(a)(1).
The Department next argues its perception of the legislative intent is evidenced by the statutory provisions prohibiting the Department from suspending a license, and the ALJ from sustaining a departmental license suspension, if the arrested person’s subsequent breath test indicates an alcohol concentration of below 0.10 “at the time the specimen is taken.” Id. §§ 524.012(c)(1), 524.035(d)(1). We again disagree. At most, these provisions reflect a legislative assumption that the arrested person’s alcohol concentration at the time the specimen is taken will be higher than his alcohol concentration at the time of the stop, i.e., he was still absorbing, not eliminating, alcohol at the time of the stop. If this assumption were applied to persons whose breath tests showed alcohol concentrations of 0.10 or more at the time the specimens were taken, we would be required to presume not that Míreles’ alcohol concentration at the time of the stop was 0.16 or above — the presumption the Department argues — but that it was below 0.16. This provision thus suggests a presumption that is precisely the opposite of the presumption the Department seeks.
Finally, the Department argues its perception of the legislative intent is demonstrated by the statutory provisions for proving and challenging the reliability of the breath test machine and the validity of the breath test results. See id. §§ 524.038-.039. However, these provisions do not remotely address the Department’s burden to prove at the administrative hearing the arrested person’s alcohol concentration was 0.10 or more while driving and, to the extent they might have any bearing on the issue otherwise, they would appear to evidence a legislative intent that *443the Department prove its ease without the benefit of statutory presumptions.
In sum, the Texas administrative license suspension statutes do not expressly create a mandatory or rebuttable presumption that an arrested person’s alcohol concentration at the time a breath specimen is taken is the same or lower than his alcohol concentration at the time of the stop. And the Department’s effort to find such a presumption in isolated provisions is not only unpersuasive but directly contrary to the legislative intent embodied in the express terms of the statutes, which require the Department to prove by a preponderance of the evidence, and require the ALJ to affirmatively find, the arrested person’s alcohol concentration was 0.10 or more while driving. Accordingly, neither the statutes nor the statutory framework fill the evidentiary void.
Conclusion
Applying the principles governing legal sufficiency analysis established by the Supreme Court of Texas to the record in this case does not reveal even a “mere scintilla” of evidence to support the ALJ’s finding that Míreles “was operating a motor vehicle in a public place ... with an alcohol concentration of 0.10 grams or greater of alcohol per 210 liters of breath.” And this evidentiary void is not filled by an express or implied statutory presumption in Texas’ administrative license suspension statutes or the statutory framework. Accordingly, we reverse the trial court’s judgment and render judgment reinstating Míreles’ license. However, we take this step with the greatest reluctance.
Like so many others, we are keenly and personally aware of “the tremendous toll of death, injury, and grief caused by those who, under the influence of alcohol or drugs, drive steel juggernauts capable of high speeds and devastating destruction.” McLean, 963 F.2d at 1307. But neither the tragedy inflicted by impaired drivers nor our sensitivity to it reheves us of the promise we made to follow the law established by the Texas Legislature and the higher courts. We can only hope the legislature and the Department find a sufficient means to define and prove the facts necessary to ensure the licenses of impaired drivers continue to be suspended automatically.5
. Texas courts, as well as modern commentators, have applied this scope of review to legal sufficiency review generally; Chief Justice Calvert, on the other hand, stated this scope of review applies only within the context of "scintilla review.” Calvert, 38 Tex. L.Rev. at 364. Although this distinction might be important in another case, it is immaterial in this one because only scintilla review is involved.
. We approach the science evidence regarding the absorption and elimination of alcohol cautiously, in light of our collective lack of scientific training and our reluctance to go outside the record to conduct independent research. However, we know of no way to evaluate the validity of the ALJ’s inference other than to evaluate the Department’s premise that a person’s alcohol concentration at the time breath specimens are submitted is "some evidence” of the person's alcohol concentration one hour earlier. See Merrell Dow Pharm., Inc. v. Havner, 953 S.W.2d 706, 711-12 (Tex.1997) (unreliable scientific testimony admitted without objection is "no evidence"), cert. denied, — U.S. -, 118 S.Ct. 1799, 140 L.Ed.2d 939 (1998); cf. Emerson v. State, 880 S.W.2d 759, 764-65 (Tex.Crim. App.) (holding court is authorized to take judicial notice of scientific and legislative facts necessary to decide the reliability of the horizontal gaze nystagmus test to determine admissibility of expert testimony regarding the test results), cert. denied, 513 U.S. 931, 115 S.Ct. 323, 130 L.Ed.2d 284 (1994). But cf. Maritime Overseas Corp. v. Ellis, 971 S.W.2d 402, 409 (Tex.1998) (requiring objection to unreliable scientific testimony), cert. denied, - U.S. -, 119 S.Ct. 541, 142 L.Ed.2d 450, 67 USLW 3188 (1998). And the alternative is to either accept or reject the Department’s premise because other courts have done so or to reject the premise because the Department, as the party with the burden of proof, failed to prove it with expert testimony in this case. Neither alternative would appear to yield a satisfactory resolution of the issue presented and, in any event, it appears the science underlying the absorption and elimination of alcohol, at least to the extent involved in this case, is well-settled.
. Indeed, if the record is silent on the factors affecting absorption and elimination, it is questionable whether retrograde extrapolation testimony would amount to "some evidence.” See, e.g., Merrell Dow Pharm., 953 S.W.2d at 712 (expert testimony that is based upon "possibility, speculation, and surmise,” rather than a reliable scientific basis, is "no evidence”).
. See, e.g., Finney v. State, 686 N.E.2d 133, 135 (Ind.Ct.App.1997), trans. denied; State v. Korhn, 41 Conn.App. 874, 678 A.2d 492, 493 n. 3, cert. denied, 239 Conn. 910, 682 A.2d 1010 (1996); State v. Taylor, 132 N.H. 314, 566 A.2d 172, 174 (1989); Ransford v. District of Columbia, 583 A.2d 186, 189 (D.C.1990).
. See., e.g., Settani v. Commissioner of Motor Vehicles, 48 Conn.App. 418, 710 A.2d 816, 817 (1998) (1994 amendment to Pennsylvania license suspension statute ''eliminated the requirement that the hearing officer make a finding of the level of the operator’s BAC at the time of operation and allowed the commissioner to suspend the license if the breath tests were administered within two hours of the time of operation and revealed an illegal level at the time of testing”), cert. denied, 245 Conn. 915, 719 A.2d 1167 (1998); Fitzgerald & Hume, 66 Mass. L.Rev. at 36 (recommending a breath sample within ten to fifteen minutes after stop and another approximately one-half hour after the first).