Alcohol use offers social benefits for young adults, but also carries the risk of significant negative consequences. Better understanding of processes driving alcohol use in this population can prevent these harms. Young adults have drinking patterns in common due to shared life circumstances (e.g., moving out, going to college) as well as patterns unique to each individual. Evidence for these processes have been limited by the use of methods that fail to capture the complex, heterogenous, multivariate nature of drinking. In this study, we overcome these limitations with a computational modeling approach called group iterative multiple model estimation (GIMME). We studied a sample of 97 young adults who completed daily surveys for up to 120 days. With GIMME, we estimated models of each person’s unique drinking patterns by searching all possible dynamic relations among self-reported alcohol consumption and various cognitive, motivational, and emotional constructs. This method allowed us to identify common and uncommon drinking processes in a bottom-up, data-driven manner. We found a clear normative pattern of drinking: young adults drink more per occasion when they expect positive outcomes and are motivated to have fun and enhance social experiences, which leads to positive consequences and–less frequently–negative consequences. These results suggest that public health initiatives to reduce harmful young adult drinking may have the widest impact by promoting non-drinking, rewarding social activities. We also identified young adults with uncommon, and potentially problematic, alcohol use patterns who could benefit from interventions tailored to their unique drinking processes.