This article presents a novel method for measuring the acute effects of alcohol. One hundred twenty nonproblem drinkers aged 21–28 participated in 3 alcohol administration sessions that produced peak blood alcohol concentrations (BACs) near .09 g%. Subjective intoxication ratings were taken at multiple points across rising and falling BACs. Mathematical modeling techniques decomposed intoxication ratings into a tonic component sensitive to BAC level and a phasic component sensitive to BAC rate of change. This model provided a good fit to observed data. Tonic and phasic gain parameters showed high repeatability across sessions. The average phasic gain parameter was about 4 times larger than the average tonic gain parameter, indicating that subjective intoxication is usually more affected by BAC change than by BAC level. The associations of drinking practices with tonic and phasic gain parameters varied by gender and family history of alcoholism. Tonic-phasic modeling allows individual and group differences in the acute effects of alcohol to be studied as time-dynamic processes.