Significance of variations in blood: breath partition coefficient of alcohol

A helium-neon laser was used to measure the alcohol content of breath from six volunteers at regular intervals over up to four hours. The corresponding blood values were calculated with a blood : breath partition coefficient of 2100. When these values were compared with those obtained by direct measurement it was obvious that substantial variations occurred from one person to another in the derived values and that even in the same person the use of the partition coefficient of 2100 led to significant differences between the direct and derived values for blood, and these differences changed with time. Thus the assertion that a constant partition coefficient of 2100 exists between alcohol in blood and that in breath is not supported by the evidence. Accordingly the use of such a partition coefficient to derive blood alcohol values for law enforcement is not justified.