The inter- and intra-individual components of variation in the saliva/blood alcohol ratio have been calculated from experiments with 48 male subjects after they drank 0.72 g of ethanol per kilogram of body weight as neat whisky after a short fast. Saliva and blood ethanol profiles were monitored at 30--60 min intervals for up to 7 h after intake. The analytical component of variation inherent in an estimate of the saliva/blood alcohol ratio, expressed as coefficient of variation, was 1.75%. I calculated saliva/blood ethanol ratios for each subject at each sampling time by taking the antilogarithm of the difference (log saliva alcohol--log blood alcohol). The mean ratio between 60 and 360 min after drinking was 1.077 (n = 336) with 95% confidence limits of 1.065 and 1.088. Moreover, the individual ratios showed no systematic variation throughout the absorption, distribution, and elimination phases of ethanol metabolism. Using a two-way analysis of variance and allowing for analytical sources of variation, I determined that the inter- and intra-subject variance components were 53 and 47% of the total biological variation. The saliva/blood alcohol ratio during ethanol metabolism, determined once in a single individual, had a biologically derived coefficient of variation of 10%.