Experimental subjects drank known amounts of vodka diluted with carbonated lemonade, and breath testing commenced after allowing for absorption and equilibration of the alcohol throughout the water compartments of the body. Measurements of breath expiration and temperature, as well as breath alcohol readings, were carried out consecutively over a period of 8 minutes, each at an increased volume of breath before the sample for analysis was taken. Results showed that lower breath pressures stemming from resistance to breathing, resulting from placement of alcohol test instruments, did not cause variation in breath alcohol concentrations when the volume of breath expired before analysis was kept constant. An analysis of the relationship between breath temperature and volume of air expirated showed a linear increase in temperature after a greater volume of air had been expirated. Also, higher readings of alcohol concentration were obtained after an increased volume of air had been released by experimental subjects. It is concluded that breath temperature is more important than expired breath volume to the time of sampling when estimating blood alcohol concentrations. Therefore, temperature differences could be sufficient to explain variations in the readings of the test instruments used in the study. It is suggested that future breath-testing instruments could be made more accurate by incorporating a temperature-sensing device. One reference, four figures, and four tables are provided.