The authors examine both the alcohol consumption pattern of freshmen students during their first semester and the degree to which social modeling of peer behavior impacts consumption. A total of 534 students, residing on campus, were prospectively examined at four 30-day intervals. Data were evaluated on the basis of age, gender, and the effects of time using generalized estimating equations (GEEs). Results reflected nonsignificant increases in the amount of alcohol consumed; however, affiliations with alcohol-consuming peer groups was significantly associated with increased alcohol consumption. The mean number of drinks consumed in the past 30 days remained the same for whites (x[superscript 2] = 3.35, 3 df, P = 0.3411) but increased slightly for blacks (x[superscript 2] = 7.99, 3 df, P = 0.0462). Prevention programs should include screening for growth in alcohol consumption among first-year university students, and such screenings should include the extent and nature of affiliation with peers who consume alcohol. (Contains 2 figures and 2 tables.)