Snap Shot
- The question: What do breath, whole-blood, and hospital serum or plasma alcohol results each measure, how do they relate to one another, and what should a court understand before treating any of them as a person’s blood alcohol concentration (BAC)?
- What the research shows: Evidential breath instruments measure alcohol in exhaled breath and report it on a scale that assumes a fixed blood-to-breath ratio of 2100:1. In one study of 799 Swedish drivers the average measured ratio was about 2400:1 (2407:1), with 95% of individuals between about 1981 and 2833 (Jones & Andersson, 1996), and ratios vary from person to person, so breath results are generally somewhat lower than simultaneous venous blood results, though not for every person or at every point on the curve. Hospital laboratories usually test serum or plasma, which contains more water than whole blood and therefore more alcohol; serum results average about 14% to 15% higher than whole-blood results, with individual ratios spanning roughly 0.9 to 1.6.
- What it does not show: No single conversion makes a breath or serum result exactly equal to a particular person’s whole-blood BAC; each conversion carries a range. A correctly measured result also says nothing by itself about what the person remembered, how the person appeared, or whether the person could consent.
- The law: Breath tests incident to arrest are permitted without a warrant, but blood tests generally require a warrant, consent, or exigency (Birchfield v. North Dakota, 2016; Missouri v. McNeely, 2013; Mil. R. Evid. 312(d)). Forensic laboratory reports are subject to the Confrontation Clause (Melendez-Diaz, 2009; Bullcoming, 2011; Smith v. Arizona, 2024). Arizona prescribes foundation for breath results (A.R.S. § 28-1323) and, when an officer has probable cause to believe a DUI was committed, entitles the officer to part of a blood sample drawn for any reason (A.R.S. § 28-1388(E)); consent given only in response to an admonition that the law requires testing does not by itself show voluntary consent (State v. Valenzuela, 2016).
- Issues for each side: Either side can legitimately rely on properly obtained breath, blood, or hospital results; both should identify exactly what specimen was tested, by what method and when, convert serum results with a stated range, and avoid presenting a converted or breath value as a direct measurement of whole-blood BAC.
Alcohol evidence reaches court in several forms. A police breath instrument produces a printed result; a crime laboratory reports alcohol in whole blood; and in many sexual assault, assault, and injury cases the only number comes from a hospital, where a clinical laboratory tested serum or plasma for medical purposes. These results are often treated as interchangeable, and legal limits are written as if they were. They are not identical. Each measures a different specimen, each relates to whole-blood BAC through a ratio that varies between people, and each has its own requirements for timing, handling, and proof.
This paper explains what breath instruments measure and the 2100:1 convention, how much the actual blood-to-breath ratio varies, how serum and plasma differ from whole blood and how hospital results are converted, and the timing, custody, and confrontation issues that attach to each. It complements “Alcohol Toxicology for Attorneys” and Retrograde Extrapolation and the Limits of the Widmark Formula, which address absorption, elimination, and back-calculation. Toxicology for drugs other than alcohol is addressed in Counterintuitive Testimony: Drug-Facilitated Sexual Assault: Toxicology, Detection Limits, and What the Results Can Show.
What Breath Instruments Measure
When blood passes through the lungs, alcohol moves from the blood into the air in the tiny air sacs, and exhaled breath carries it out. Evidential breath instruments measure the alcohol in a sample of end-expired breath, either by infrared spectrometry or with an electrochemical fuel cell (Jones, 2019). In the United States the result is reported in grams of alcohol per 210 liters of breath, a unit chosen so that a breath result reads the same as a blood result in grams per 100 milliliters if the blood contains 2,100 times as much alcohol as an equal volume of breath. Military law and Arizona law both define the legal limit for breath directly in these units (Art. 113(b)(3), UCMJ; A.R.S. § 28-101, defining alcohol concentration), so a breath result above the limit satisfies a breath-based element without any conversion. The ratio matters when a breath result is used to infer a blood concentration, for example to compare it with a later blood test, to estimate impairment, or to apply research expressed in blood units.
Countries differ in the ratio they build into their instruments: the United States and Canada use 2100:1, Britain 2300:1, and much of Europe 2000:1 (Jones, 2008). Those differences reflect conventions adopted for law enforcement rather than a single physiological constant.
How Much Does the Blood-to-Breath Ratio Vary?
The most cited data come from 799 Swedish drivers who gave a breath test on an infrared instrument and then a venous blood sample about 30 minutes later. After correcting the blood result for alcohol eliminated between the two samples, the average blood-to-breath ratio was about 2407:1, with 95% of individuals falling between about 1981 and 2833; the ratio did not depend meaningfully on age, sex, or BAC (Jones & Andersson, 1996). Because the average real ratio exceeds 2100, a breath result read on the 2100 scale generally understates the simultaneous venous blood concentration, by roughly 13% on average. But the variation means the relationship is not uniform: after correction, about 4% of the drivers had ratios below 2100, for whom the breath result would overstate venous blood alcohol.
The phase of drinking also matters. While alcohol is still being absorbed, arterial blood (which supplies the lungs’ exchange with breath) carries more alcohol than venous blood drawn from the arm. In a study in which 13 men received alcohol by intravenous infusion, arterial concentrations exceeded venous concentrations by about 0.010 g/dL during the infusion, breath results tracked arterial blood more closely, and the differences largely disappeared within minutes after the infusion ended (Jones et al., 1997). A forensic reference summarizing several studies describes the venous blood-to-breath ratio as lower during absorption and higher later in the elimination phase (Jones, 2008). A breath test taken soon after drinking may therefore read higher relative to a venous blood sample than one taken hours later.
Physiologists have also questioned the model on which breath testing rests. A review of gas-exchange research concluded that alcohol is exchanged with the lining of the airways during exhalation, so that exhaled breath does not exactly reflect alveolar air, and that the size of this effect depends on a person’s lung volume, breathing pattern, and other characteristics, creating potential biases for some individuals (Hlastala & Anderson, 2016). The practical significance of these effects for forensic breath results is debated, and this paper does not resolve the dispute. Residual alcohol in the mouth from a recent drink, regurgitation, or certain dental conditions can also raise a breath reading; observation or deprivation periods and duplicate tests are used to guard against it, and one reference recommends waiting at least 15 minutes after the last drink (Jones, 2008). Body temperature can also matter: in a laboratory study that raised core temperature by about 2.5 degrees Celsius in intoxicated subjects with a warm-water bath, breath alcohol rose about 8.6% relative to blood alcohol for each degree Celsius, to a maximum of about 23% (Fox & Hayward, 1989); the abstract does not report the number of subjects, and the study does not show how the finding applies to an ordinary fever. Arizona’s breath-test statute includes duplicate tests within 0.02 or a twenty-minute observation period among its foundational requirements (A.R.S. § 28-1323(A)). The Arizona Department of Public Safety’s rules separately define a deprivation period for duplicate breath tests as at least 15 minutes immediately before the test, during which the subject may not drink, eat, vomit, smoke, or put anything in the mouth (Ariz. Admin. Code R13-10-101), and require duplicate results within 0.020 of each other (Ariz. Admin. Code R13-10-104). The 15-minute deprivation period and the statute’s 20-minute observation alternative are different provisions; as of November 2026, counsel should confirm which one a given record invokes.
Whole Blood Versus Serum or Plasma
Whole blood consists of cells suspended in plasma; serum is the liquid that remains after blood clots, and plasma is the liquid portion of blood kept from clotting. Because alcohol distributes according to water content, and serum and plasma contain more water than whole blood, the same blood yields a higher alcohol concentration in serum or plasma than in whole blood. Crime laboratories and legal limits generally use whole blood; hospital clinical laboratories generally test serum or plasma.
Two studies frame the conversion. In 235 subjects whose serum and whole blood were tested by gas chromatography, serum-to-whole-blood ratios ranged from 1.04 to 1.26, with a mean of 1.14 (Charlebois et al., 1996). In 211 hospital patients, the ratios ranged more widely, from 0.88 to 1.59, with a median of 1.15; the central 95% of patients fell between about 0.95 and 1.40, and the ratio did not depend on the alcohol level or the hematocrit (Rainey, 1993). Dividing a serum result by about 1.14 or 1.15 therefore gives a reasonable central estimate of whole-blood BAC, but an individual’s true ratio may differ, and the range matters near a legal threshold.
Arithmetic illustrates the point. A hospital serum result of 150 mg/dL corresponds to 0.150 g/dL in serum. Divided by 1.14, it suggests a whole-blood BAC of about 0.13 g/dL. Applying the central 95% range from Rainey (1993), the whole-blood value could plausibly lie between about 0.107 and 0.158. The serum figure itself, presented without conversion, would overstate the whole-blood concentration for most people.
Hospital Records: Timing, Method, and Units
Hospital results raise questions beyond the serum conversion. Clinical laboratories test alcohol to guide treatment, not to produce evidence, and their methods, quality controls, documentation, and specimen handling may differ from those of a forensic laboratory; gas chromatography is the forensic reference method, and other methods are also in use (Jones, 2019). The record should therefore be read for the specimen type, the method, the units (milligrams per deciliter, grams per liter, or millimoles per liter), and the exact time of collection, which may differ from the time the result was reported. The time of collection is the starting point for any extrapolation, and in injury cases a person may arrive at a hospital hours after the events, after receiving treatment, or after additional drinking. Chain of custody for a hospital specimen is usually less formal than for an evidential sample, and a portion of the original specimen may or may not have been retained.
Hospital results can also be the best evidence available. In many sexual assault cases no police breath or blood test is taken, and a sample drawn for clinical care may be the only contemporaneous measurement of a complainant’s or an accused’s alcohol level. Properly interpreted, with the specimen identified and the conversion stated as a range, it can anchor a timeline; the paper Reconstructing a Blackout Period from Digital and Documentary Evidence discusses how such records fit with others.
Case Example. In Bullcoming v. New Mexico (2011), a driver was charged with aggravated driving while intoxicated after his vehicle rear-ended a pickup truck, and a state laboratory analyst certified that his blood sample, tested by gas chromatography, showed a BAC of 0.21 g/dL. The analyst who performed the test did not testify; instead, the prosecution called a different analyst who was familiar with the laboratory’s procedures but had neither performed nor observed the test. The Supreme Court held that this surrogate testimony did not satisfy the Confrontation Clause, because the certified report was testimonial and the accused was entitled to confront the analyst who made the certification, unless that analyst was unavailable and the accused had a prior opportunity for cross-examination. The case illustrates that a BAC result, however reliable the instrument, reaches the factfinder through people whose work can be examined.
Limits of the Evidence
Most blood-to-breath data come from Scandinavian drivers tested at high concentrations on particular instruments, and ratios may differ with other instruments, populations, and phases of drinking. Serum-to-blood studies used a few hundred subjects each, and the wider range in hospital patients may reflect illness, treatment, or analytic differences. Studies of arterial and venous differences used small samples of men, sometimes with intravenous rather than oral alcohol. Sex was examined in the largest breath-ratio study and was not associated with the ratio (Jones & Andersson, 1996); several of the smaller studies did not analyze it. The research describes how specimens relate on average and across ranges; it cannot determine an individual’s true ratio after the fact, and none of it addresses memory, behavior, or capacity.
Questions to Ask About This Research
When a breath result is converted to a blood concentration, counsel can ask what ratio was used, whether the conversion accounts for the range of individual ratios (Jones & Andersson, 1996), and whether the person may still have been absorbing alcohol at the time of the test (Jones et al., 1997). When a hospital result is offered, counsel can ask whether it was serum, plasma, or whole blood, what method and units the laboratory used, when the specimen was collected rather than reported, and what conversion factor and range were applied (Charlebois et al., 1996; Rainey, 1993). When an expert relies on the 2100:1 convention as favorable to the person tested, counsel can ask whether that holds for this person at this phase of drinking, and when an expert disputes the breath result on physiological grounds, counsel can ask what evidence shows the claimed effect in this person rather than in principle (Hlastala & Anderson, 2016). These questions are explained in “Reading Research Claims: A Guide for Attorneys” on this site.
Breath and Blood Testing and the Law
Federal and Military Standards
The Fourth Amendment governs how samples are obtained. A compelled blood draw is a search; Schmerber v. California (1966) upheld a warrantless draw on probable cause where delay threatened the loss of evidence, and Missouri v. McNeely (2013) held that the natural dissipation of alcohol does not create a per se exigency in every case. Birchfield v. North Dakota (2016) held that breath tests, but not blood tests, may be administered without a warrant as a search incident to a lawful arrest for drunk driving, and that a motorist may not be criminally punished for refusing a warrantless blood test. Mitchell v. Wisconsin (2019), in a plurality opinion, concluded that when a driver’s unconsciousness or stupor requires a trip to a hospital before a breath test can be given, the exigent-circumstances rule almost always permits a warrantless blood test, and it remanded to let the driver try to show that his case was unusual. In the military, nonconsensual extraction of body fluids is admissible if obtained under a search warrant or authorization, or without one only when probable cause existed and the delay needed to obtain authorization could have resulted in destruction of the evidence, and the extraction must be performed reasonably by a person with appropriate medical qualifications (Mil. R. Evid. 312(d)). Article 113 sets limits for both blood (per 100 milliliters) and breath (per 210 liters) as shown by chemical analysis.
Laboratory reports are subject to confrontation. Sworn certificates of forensic analysis are testimonial (Melendez-Diaz v. Massachusetts, 2009), a certified blood-alcohol report cannot be introduced through a surrogate analyst who neither performed nor observed the test (Bullcoming v. New Mexico, 2011), and when an expert conveys an absent analyst’s statements in support of the expert’s opinion and the statements support it only if true, they are offered for their truth (Smith v. Arizona, 2024); the Court left for remand whether the statements in that case were testimonial. The military decisions on laboratory reports, including the two United States v. Blazier opinions (2010) and United States v. Sweeney (2011), are discussed in the Counterintuitive Testimony paper on drug-facilitated sexual assault. Expert testimony on conversions and ratios is governed by Federal Rule of Evidence 702, as amended December 1, 2023, Daubert v. Merrell Dow Pharmaceuticals, Inc. (1993), and Kumho Tire Co. v. Carmichael (1999); Military Rule of Evidence 702 tracks the federal rule, and the 2023 amendment took effect in the military rule in 2025 (Joint Service Committee on Military Justice, 2025); practitioners should confirm current text. Military judges also apply United States v. Houser (1993).
Arizona
Under Arizona’s implied-consent statute, a person who operates a motor vehicle in the state gives consent to tests of blood, breath, urine, or other bodily substance for alcohol concentration or drug content if arrested for an offense arising from acts alleged to violate the state’s driving-under-the-influence provisions (A.R.S. § 28-1321(A)). In State v. Valenzuela (2016), the Arizona Supreme Court held that consent given only in response to an admonition that Arizona law “requires” submission is not shown to be freely and voluntarily given, though it applied the good-faith exception to the arrest at issue. A breath result is admissible on a showing that an approved device was used by an operator holding a valid permit, that duplicate tests agreed within 0.02 or the operator observed the person for the twenty minutes before the test, that an approved operational checklist was followed, and that the device was in proper working order (A.R.S. § 28-1323(A)). When blood is drawn for any reason, including medical treatment, and an officer has probable cause to believe a DUI was committed, a portion of the sample must be provided to the officer on request (A.R.S. § 28-1388(E)); persons tested must be given a reasonable opportunity to arrange an independent test (A.R.S. § 28-1388(C)). Smith v. Arizona (2024) arose from an Arizona laboratory analysis and rejected the basis-evidence reasoning Arizona courts had applied. Counsel should research current Arizona authority on the conversion of hospital serum results to whole-blood equivalents. Expert testimony is governed by Arizona Rule of Evidence 702, amended effective January 1, 2024 (Ariz. Sup. Ct. Order R-23-0004), and State v. Salazar-Mercado (2014).
Special Issues in Legal Practice
The principles described above apply with equal force to every party in a legal proceeding. The science operates the same way regardless of who is asking the questions, and the research summarized here does not by itself favor either side. What differs between the parties is not the science but the vantage point — the particular places where these issues tend to surface, and the particular errors each party is best positioned to guard against. The two summaries below identify those recurring issues. They are offered for orientation, not as litigation strategy; in every case the controlling question is what the evidence in that case will, and will not, support, and the expert’s role is to inform the trier of fact rather than to advocate for either side.
Special Issues for Prosecutors
The research supports several propositions prosecutors can legitimately present. Because the average blood-to-breath ratio exceeds 2100:1, a breath result generally understates the simultaneous venous blood concentration for most people (Jones & Andersson, 1996), and breath limits defined in breath units apply without conversion. A hospital serum result, properly converted with a stated range, can be reliable contemporaneous evidence of a complainant’s or accused’s alcohol level when no police test was taken. Calling the analyst who performed the test, documenting collection times and specimen type, and obtaining hospital samples through the statutory and constitutional routes reduce later disputes.
The research also counsels caution. Presenting a serum figure as if it were whole-blood BAC overstates the concentration for most people, and a single conversion factor without a range can matter near a threshold. The 2100 convention does not favor every individual, particularly during absorption, and testimony that a breath result is always conservative goes beyond the data. Hospital methods, custody, and documentation may not match forensic standards, and Bullcoming and Smith limit how a laboratory’s work can be presented through a substitute witness.
Special Issues for Defense
The research supports several propositions defense counsel can legitimately present. Breath results depend on a ratio that varies between people and with the phase of drinking, and during absorption breath can read higher relative to venous blood (Jones et al., 1997). Serum and plasma results must be converted, and the individual ratio may range widely (Rainey, 1993). The specimen type, method, units, timing, and custody of hospital results are legitimate subjects of inquiry, as are the foundational requirements for breath results, the warrant and consent rules for blood, and the right to confront the analyst whose work is offered.
The research also counsels caution about the limits that apply to the defense. For most people the 2100 convention makes breath results lower, not higher, than venous blood, and a general claim that breath testing overstates BAC is not supported for the average person; physiological critiques of breath testing describe potential biases whose size in a particular person is usually unknown. Serum conversion lowers but does not eliminate an elevated result, and an extremely high hospital value remains high after conversion. The same specimen and conversion questions apply when a defense theory relies on a complainant’s or an accused’s hospital result to show heavy intoxication.
This paper is for educational purposes only and is not legal advice. The author is a psychologist, not an attorney; summaries of law are general and must be verified against current authority or with a licensed attorney.
References
Charlebois, R. C., Corbett, M. R., & Wigmore, J. G. (1996). Comparison of ethanol concentrations in blood, serum, and blood cells for forensic application. Journal of Analytical Toxicology, 20(3), 171–178. https://doi.org/10.1093/jat/20.3.171
Fox, G. R., & Hayward, J. S. (1989). Effect of hyperthermia on breath-alcohol analysis. Journal of Forensic Sciences, 34(4), 836–841. https://doi.org/10.1520/JFS12711J
Hlastala, M. P., & Anderson, J. C. (2016). Alcohol breath test: Gas exchange issues. Journal of Applied Physiology, 121(2), 367–375. https://doi.org/10.1152/japplphysiol.00548.2015
Joint Service Committee on Military Justice. (2025). Amendments to the Military Rules of Evidence since 1984 (G. E. Maggs, Ed.; 2025 ed.). https://jsc.defense.gov/Portals/99/MRE%20Amendments%202025.pdf
Jones, A. W. (2008). Biochemical and physiological research on the disposition and fate of ethanol in the body. In J. C. Garriott (Ed.), Garriott’s medicolegal aspects of alcohol (5th ed., pp. 47–156). Lawyers & Judges Publishing.
Jones, A. W. (2019). Alcohol, its analysis in blood and breath for forensic purposes, impairment effects, and acute toxicity. WIREs Forensic Science, 1(6), Article e1353. https://doi.org/10.1002/wfs2.1353
Jones, A. W., & Andersson, L. (1996). Variability of the blood/breath alcohol ratio in drinking drivers. Journal of Forensic Sciences, 41(6), 916–921. https://doi.org/10.1520/JFS14025J
Jones, A. W., Norberg, Å., & Hahn, R. G. (1997). Concentration-time profiles of ethanol in arterial and venous blood and end-expired breath during and after intravenous infusion. Journal of Forensic Sciences, 42(6), 1088–1094. https://doi.org/10.1520/JFS14265J
Rainey, P. M. (1993). Relation between serum and whole-blood ethanol concentrations. Clinical Chemistry, 39(11), 2288–2292.
Legal Authorities
Birchfield v. North Dakota, 579 U.S. 438 (2016).
Bullcoming v. New Mexico, 564 U.S. 647 (2011).
Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993).
Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999).
Melendez-Diaz v. Massachusetts, 557 U.S. 305 (2009).
Missouri v. McNeely, 569 U.S. 141 (2013).
Mitchell v. Wisconsin, 588 U.S. 840 (2019).
Schmerber v. California, 384 U.S. 757 (1966).
Smith v. Arizona, 602 U.S. 779 (2024).
State v. Salazar-Mercado, 234 Ariz. 590, 325 P.3d 996 (2014).
State v. Valenzuela, 239 Ariz. 299, 371 P.3d 627 (2016).
United States v. Blazier, 68 M.J. 439 (C.A.A.F. 2010).
United States v. Blazier, 69 M.J. 218 (C.A.A.F. 2010).
United States v. Houser, 36 M.J. 392 (C.M.A. 1993).
United States v. Sweeney, 70 M.J. 296 (C.A.A.F. 2011).
U.S. Const. amends. IV, VI.
10 U.S.C. § 913 (Article 113, Uniform Code of Military Justice).
Ariz. Admin. Code R13-10-101, R13-10-104.
Ariz. Rev. Stat. Ann. § 28-101(2).
Ariz. Rev. Stat. Ann. § 28-1321(A).
Ariz. Rev. Stat. Ann. § 28-1323(A).
Ariz. Rev. Stat. Ann. § 28-1388(C), (E).
Fed. R. Evid. 702.
Mil. R. Evid. 312(d), 702, 1102.
Ariz. R. Evid. 702.
Ariz. Sup. Ct. Order R-23-0004 (amending Ariz. R. Evid. 702, effective Jan. 1, 2024).
Related topic papers: Alcohol Toxicology for Attorneys: Absorption, Distribution, Elimination, and Stages of Intoxication · Retrograde Extrapolation and the Limits of the Widmark Formula · Reconstructing a Blackout Period from Digital and Documentary Evidence · Blackouts and Blood Alcohol Concentration
Related topics from Arizona Forensic Services: Counterintuitive Testimony: Drug-Facilitated Sexual Assault: Toxicology, Detection Limits, and What the Results Can Show · Courts-Martial