Objective: The present study was designed to test for projection, conformity, or reciprocal associations among attitudes, injunctive norms, and drinking. Assuming that these constructs are not independent, we proposed three possible trajectories. A conformity model would suggest that injunctive norms should temporally precede drinking or attitudes. Alternatively, a projection model would suggest that attitudes or alcohol use would precede injunctive norms. Furthermore, by examining the processes over three time points, the current study would also examine whether both conformity and projection processes were at work, indicating reciprocal relationships. Method: Participants included 249 college students (63.1% female), who participated as a control group in a larger intervention trial. Structural equation modeling was used to examine cross-sectional and prospective associations among injunctive norms, attitudes, and drinking across each of the three time points. Results: Findings demonstrated three significant cross-lagged associations. Injunctive norms at Time 1 was significantly associated with drinking at Time 2 (conformity), and both attitudes and drinking at Time 2 were significantly associated with injunctive norms at Time 3 (projection). The pattern of cross-lagged associations suggested one meaningful indirect pathway, from Time 1 injunctive norms to Time 2 drinking to Time 3 injunctive norms (reciprocal association). Conclusions: The present study suggests that both the conformity and the projection processes seem to be important and evident for college student drinking when considering injunctive norms and drinking over time. Interventions that focus on both conformity and projection may be particularly effective at reducing longer-term alcohol use.