The participants were 24 healthy male volunteers who ranged in age from 22 to 56 years. The participants took part in a 3-hour social drinking session. A blood sample for alcohol analysis was taken from each participant approximately 1 hour after drinking stopped; another was taken approximately 3.5 hours after drinking stop. The estimates of BACs at the blood sampling time points were made assuming that each person had a constant blood alcohol clearance rate in the range of 10 to 20 mg/dL/h (0.01 to 0.02 g/dL/h) over the entire experimental period. A variety of methods were used to estimate the volume of distribution for alcohol. All BAC estimations were made assuming complete absorption and full equilibration of the total alcohol dose. Actual BACs were usually within or very close to the range of forward estimates based on the known alcohol doses. Furthermore, most BACs measured about 1 hour after cessation of drinking were within or very close to the predicted range based on back extrapolation from the actual 3.5-hour BAC result. Findings supported the opinion that back and forward estimation of BAC can seldom be performed with great accuracy. However, findings also suggest that such estimates are unlikely to be subject to gross errors if the person has been in a normal social drinking situation and certain assumptions are used. Different assumptions might be more useful if the person is believed to be a chronic heavy drinker. In every case, the results given by the formula must be interpreted according to the circumstances of the drinking situation. Tables, figures, and 29 references (Author abstract modified)