Schaffer Online Library of Drug Policy Sign the Resolution
Contents | Feedback | Search
DRCNet Home
| Join DRCNet
DRCNet Library | Schaffer Library | Drugs and Driving 


Breath Alcohol Testing and Daubert Criteria: The Jurisprudence of Science in DWI Cases

R. T. Kennedy, JD

State of New Mexico, Bernalillo County Metropolitan Court, P.O. Box 133, Albuquerque, NM 87103, USA


Breath alcohol testing is fraught with inherent analytical and physiological variables which are often caught in the legal debate concerning the 'quantification of certainty'. In many jurisdictions, per se evidence of breath alcohol content (BrAC) alone is sufficient to convict a person accused of DUI. A reliable scientific process and an evenly-applied judicial standard can fairly resolve unnecessary litigation and expense.

In the United States, the case of Daubert v. Merrill Dow Pharmaceuticals has interjected scientific concepts of demonstrable error rates and validity to drunk driving litigation. Together with existing research concerning the analytical variables inherent and correlation of error rates with reasonable to breath testing, a review of the literature supports the ongoing development of valid testing techniques together with a jurisprudential method of dealing with this scientific evidence.

Reflections concerning the valid testing technique doubt suggest a resolution of the scientific/judicial dilemma in the favour of acceptable reporting of breath testing results when used in the judicial context.


The jurisprudence of science changed in late 1993 when the U.S. Supreme Court decided the case of Daubert v. Merrell Dow Phamaceuticals. The Court held that the former 'general acceptance' test for the admission of scientific evidence (Frye, 1923; Throughout this paper, italicized names will refer to reported court cases) had been superseded by the more liberal Federal Rules of Evidence, and put forth a broader framework for the judicial evaluation of scientific evidence intended to more closely reflect the process of science itself. Many states have adopted the Federal Standard enunciated in Daubert as theirs. The science of breath alcohol testing will be judged by this standard. Even with this attention to scientific evidence, a parallel legal trend to eliminate many scientific considerations from DWI laws exists. This report addresses the Daubert legal framework as it applies to breath alcohol testing, and addresses per se and breath-only standards as they are adopted by states.

BrAC testing in DWI cases exists between being circumstantial evidence of intoxication and impairment, and per se evidence of a criminal violation where a BrAC level alone has been set by statute as illegal (Jones, 1987). The legislative norm increasingly tilts toward per se laws. A breath test becomes primary evidence of guilt and thus renders the legal consideration of test results to be of great import. (Dubowski, 1985) Knowing limitations of the technology, and reporting results as existing within appropriate confidence limits become crucial to the fair introduction of this increasingly important evidence (Fitzgerald and Hume, 1994). Expert and scientific testimony must assist the trier of fact to determine an issue in the case or add to its understanding of issues. Beyond this, Daubert provides courts with some framework to evaluate the utility of scientific evidence on a case-by case basis.

The evidence comes from scientific process, or is scientific in nature

Concepts of falsification of hypotheses, testability of the scientific technique and quantification of reliability all play into this evaluation. Scientific validation of the technique and methodology must be clearly demonstrable.

Breath Alcohol Testing has been utilized in DWI cases since the late 1920s. The individual analytical processes have enjoyed greater or lesser degrees of acceptance through time, but all are based upon scientific analytical principles and processes that were developed outside of the forensic context. The process by which unknown samples are tested is inextricably rooted in the scientific method.

Characteristics of certain methodologies may be criticized as lacking in scientific rigor. Forensic science has specifically been subjected to such scrutiny (Jonakait, 1994). As the Daubert case notes, "Scientific validity for one purpose is not necessarily scientific validity for other unrelated purposes." A technique for analyzing controlled samples in a laboratory may lose validity when used to analyze different samples in the field.

The underlying scientific theory should be demonstrably valid and reliable

There is no doubt breath alcohol is an indirect measure of blood alcohol which operates on the brain to cause impairment. The first scientific question can be the extent to which indirect testing can be relied upon to reliably reflect impairment of driving ability. This question has been addressed in a number of ways in the literature.

Analytical reliability of breath testing instruments has received much attention. Isolating sources of error inherent to the instrumentation, eliminating as many as possible and quantifying the rest within acceptable limits is frequently studied. The stated range of an Intoxilyzer 5000, for example as being accurate to .005% BAC or 5% with repeatability of .0042%BAC (CMI, Inc. 1994) seems to set a range of accuracy and reliability. Duplicate sampling of subjects' breath is a matter of convention; samples must generally agree within a .02% (Dubowski, 1986, Hodgson, 1995).

The scientific community recognizes that testing of human subjects' breath is fraught with many more variables than testing samples which are controlled, like the simulators used to check the analytical validity of the instruments. Blood/breath ratios, absorbtive or postabsorbtive metabolic states, body temperature, machine specificity for ethanol, diurnal cycles, hematocrit, mouth alcohol, breathing patterns and a changing notion of even where the alcohol in the breath comes from receive great attention and are recognized as being of significance when human subjects are tested for BrAC. (Hlastala,1985;Jones,1994 ; Ohlsson, 1990; Simpson, 1987 a/b; Tsu, 1988, 1991) Science continues to test, re-test and examine the the nature of breath alcohol testing. Law often requires one answer for the present case.

American courts have tended to accept breath alcohol testing as valid and reliable for purposes of admission of test results into evidence. (Trombetta, 1984; Downie, 1990) Many cases addressing breath testing in general have relied, as did the U.S. Supreme Court in Trombetta, on the certification process for breath test devices by regulatory agencies, and further pointed to a criminal defendant's ability to cross-examine or impeach the machine with evidence of his own (including scientific evidence) as a protection of procedural fairness.

The theory and its practice has been subjected to peer review

Another factor which goes to the weight of a scientific technique's reliability as evidence under Daubert is the extent to which it has been evaluated by the scientific community at large. Alcohol pharmacokinetics and the dynamics of alcohol transport in the human body have been extensively studied and subjected to peer review. The extent to which breath alcohol testing has been subjected to peer review has inspired and kept lively the debate concerning its abilities to deliver its promised product: a valid and reliable measurement of alcohol content which is relevant to a legal standard. Studies which seek to document correlations between absorptive, peak and post-absorbtive arterial and venous blood levels with simultaneous breath testing are certainly helpful (Jones, 1995), and lend scientific answers to forensic questions. Articles which establish the physiological bases for testing error complement those which establish acceptable limits of analytical error inherent to the instruments.

General acceptance of a technique and methodology within a relevant and recognized scientific community were requisite under the Frye standard which governed prior to Daubert, and is still a relevant consideration under the new paradigm. The existence of lively debate does not rule out a scientific technique from being acceptable to a court, but rather provides context within which specific issues must be resolved. Here then, also rests the distinction between admissibility of evidence and its sufficiency to meet the burden of proof which must be carried by the party offering evidence. The trier of fact remains utterly free by law to disregard perfectly admissible evidence as insufficient to support the burden of proof if it does not help consideration of the case.

The use of the theory in the case at hand was done in a manner that can be accepted as valid and reliable

Use of the methodology in the context of the case is yet another facet of Daubert analysis. Standardized procedures for the administration of breath alcohol tests in DWI cases are common, if not universal in practice. Provisions for duplicate sampling, regular calibration checks within certain limits, certification of instruments and their operators and agreement between results are widely used to establish the accuracy of the testing instrument and regulate its use. Compliance with these regulations is commonly demonstrated by keeping of detailed calibration, maintenance and test records. (Wittenbarger, 1994) More stringent requirements which are frequently used include calibration standards run between samples, and calculations of limits of confidence intervals against relevant subject populations. (Gullberg, 1995) Practices along these lines, however, are less universal. For instance, the United States Nuclear Regulatory Commission regulations require that a "failing" breath test [.04%BrAC] be confirmed by a test administered on a second instrument, and further by a GC analysis of the subject's blood upon his demand. Given the contrast between perhaps more rigidly governed testing for employment purposes and testing for purposes of furnishing evidence of drunk driving, the question of acceptable methodology will properly continue to be the subject of argument.

The technique has a known or demonstrable rate of error

The ability to state that a breath alcohol test is accurate and reliable within known limits should be important to assure the finder of fact that possible sources of error have been evaluated and accounted for. Demonstrable levels of confidence intervals and error factors in breath can be calculated. (Jones, 1994, Gullberg, 1995). The most common approach taken by the courts is that inherent margins of error of the testing device itself are a question for the fact finder in the case. Accounting for demonstrable variations in breath test accuracy caused by physiological factors present a more formidable problem. Some courts have assigned an error factor to subtract from a breath test result when converting to a blood alcohol value (Burling, 1987, McGinley, 1988). In almost diametric opposition, other courts have found that despite differences in blood-breath ratios, the danger of overestimation of a subject's alcohol content is minimal (Serrano, 1982) or merely that the "breathalyzer reads alcohol with unimpeachable accuracy"(Downie, 1990). More recently, legislative enactments defining unlawful per se alcohol levels based upon either a BAC (g EtOH/100ml blood) or BrAC (gEtOH/210l of breath) have attempted to resolve the question without resort to science. (Uniform Veh. Code 11-903) This is done for the reason that "it is observed that attacks by defendants on the partition ratio result in expensive and time consuming evidentiary hearings and undermine successful enforcement of driving under the influence laws" (California Sen. Cmte, 1989-90, Bransford 1994). Thus, per se laws no longer require that impairment enter into the jurisprudential equation, and where breath is independent evidence, relating breath alcohol to blood equivalents is found to be irrelevant. (Bransford, 1994) Defendants are left to challenging the reliability of the individual instrument, and the manner in which the test was administered, but not the underlying validity of the scientific process. (Ibid.) In states where these laws exist, it appears that Daubert will be of only limited usefulness in judging this area of scientific inquiry.


Breath alcohol testing for DWI cases is caught between the ongoing scientific dialog concerning its scientific usefulness and validity to which cases like Daubert give structure in the legal milieu, and legislative approaches which hasten to expedite the legal process to address a serious public health and safety problem. As the balance shifts from jurisdiction to jurisdiction, the value of scientific support for valid and reliable testing does not diminsh, and both views must be accommodated.


Anchorage v. Serrano, 649 P.2d 256 (AK App. 1982).

California Senate Committee on Judiciary, Analysis of Assem. Bill No 4318, 1989-90 Reg. Sess.

California v. Trombetta, 467 U.S. 479, 104 S.Ct. 2528 (U.S. Supreme Court, 1984)

CMI Inc., Intoxilyzer 5000 promotional materials, 1994

Daubert v. Merrill Dow Pharmaceuticals, 509 U.S. , 113 S.Ct. 2786 (U.S. Supreme Court, 1993)

Dubowski, K.M., "Absorbtion, Distribution and Elimination of Alcohol": Highway Safety

Fitzgerald and Hume, Intoxication Test Evidence, Civil and Criminal, NY: Clark, Boardman Callaghan 1987 (1994 supp) 4:15.1, p.164

Frye v. United States, 293 F. 1013 (U. S. Court of Appeals, D.C. Cir. 1923)

Gullberg, R., "Applying Confidence Intervals to Breath Alcohol Analyses", Abstract K1, Proceedings of the American Academy of Forensic Sciences, 1995

Hlastala, M. P. "Physiological errors associated with breath testing", The Champion 16:19, July 1985

Hodgson, B.T., "Convention for Breath Alcohol Units", Letter, Journal of Analy. Tox., 19:131 (March/April 1995)

Jonakait, R.N. "Real Science and Forensic Science", Shepards Expert and Scientific Evidence Quarterly, 1/3:435 1994

Jones, A.W., "Studies on the Physiology and Practical Application of Breath Alcohol Analysis", Lecture notes, International Association for Chemical Testing, 1994

Jones, A.W., Hahn, R.G., and Norberg, A., "Comparing the Concentrations of Ethanol in Arterial and Venous Blood and End-Expired Breath During and After Intravenous Infusion", Abstract K12, Proceedings, American Academy of Forensic Sciences, 1995.

Jones, A.W., History, Present Status and Future prospects of breath-alcohol analysis", Alcohol, Drugs and Traffic Safety, Elsevier Science Publishers, B.V. (Biomedical Div.) 1987

Ohlsson, J.; Ralph, D.D., Mandelkorn, M.A.; Babb, A.L. Hlastala, M.P., "Accurate measurement of blood alcohol concentration with isothermal rebreathing", J. Stud. Alcohol 51:6-13, 1990

People v. Bransford, et. al., 35 Cal. Reptr. 2d 613, 884 P.2d (Supreme Court of California, 1994)

Simpson, G. Accuracy and precision of Breath Alcohol Measurements for Subjects in the Absorbtive State, Clin. Chem. 33/6: 753-756 (1987) a

Simpson, G. Accuracy and precision of breath alcohol measurements for a random subject in the postabsorbtive state. Clin. Chem. 33: 261-8 (1987) b

State v. Burling, 224 Neb. 725 (Nebraska Supreme Court, 1987)

State v. Downie, 117 NJ 450, 569 A. 2d 242 (New Jersey Supreme Court, 1990)

State v. McGinley , 229 N.J. Super. 191, 550 A. 2d 1305 (New Jersey Law Division 1988)

State v. Wittenbarger, 124 Wash. 2d. 467, 880 P.2d 517 (Washington Supreme Court, 1994)

Tsu, M.E., Babb, A.L.; Ralph, D.D.; and Hlastala, M.P. , "Dynamics of heat, water and soluble gas exchange in the human airways: 1. A Model Study", Ann.Biomed. Eng., Vol. 16: 547-571, 1988 b

Tsu, M.E.; Babb, A.L.; Ralph, D.D.; and Hlastala, M.P. , Dynamics of soluble gas exchange in the airways. II. Effects of breathing conditions, Respir. Physiol. 83: 261-276, 1991 a