Prior to undertaking the design of BASS, researchers reviewed the results of earlier studies involving human subjects who had consumed alcohol to analyze the relationship of the ethanol concentration in a single breath (exhalation) as a function of time. These data were expanded upon during the current laboratory research with nondrinking human subjects in a series of experiments using individuals with a range of vital capacities and respiratory capability. The concentration of carbon dioxide as a function of exhalation time for single breath samples was monitored as each subject blew into three tubes of different flow resistance, which approximated the range of resistance offered by commercially available evidential breath testers. These data, combined with statistical data on the range of vital capacities of the U.S. driver population, serve to establish the parameters that must be controlled in order to ensure that the BASS can reproduce the extremes of single breath exhalations. Each of the essential parameters that must be controlled to produce a substitute for human breath is discussed, and the capability of the BASS to achieve that degree of control is described, together with the detailed design and operating characteristics of BASS. Data are presented that establish the need for a three-segment ethanol vapor concentration profile, in which the time and volume are controlled to produce a continuous sample that will allow the sample collection segment of breath analysis systems to be evaluated. Three specific BASS delivery cycles are recommended; these cover the range of vital capacities and flow rates that are encountered in the driving population and must be accommodated by an evidential breath tester to ensure that a sample of deep lung air is being used to analyze blood alcohol content (BAC). The report recommends that the BASS be incorporated into the National Highway Traffic Safety Administration standards for breath alcohol testing instruments to eliminate the need for human subjects to evaluate the breath-collecting capability of such instruments. Tables, diagrams, graphs, and footnotes are provided. A total of 12 references are appended. (Author abstract modified)