Latent growth mixture modeling was used to identify developmental trajectories (described in terms of demographics, exposure and resistance to a pro-drug environment, and deviant behavior) of binge drinking among 5,694 individuals who completed 6 surveys from ages 13 to 23 years: nonbingers (32%); moderate stables (37%), who had consistently low levels of bingeing; steady increasers (16%), who increased from the lowest to highest level of bingeing; adolescent bingers (9%), whose early rise in bingeing was followed by a decrease to a moderate level; and early highs (6%), who decreased from the highest level of bingeing to a moderate level. Results show considerable diversity in binge drinking patterns and the correlates of bingeing across trajectory classes.