Motives for substance use have garnered considerable attention due to the strong predictive utility of this construct, both in terms of use and problems associated with use. The current study examined the cross-lagged relations between alcohol use and motives, and marijuana use and motives over three yearly assessment periods in a large sample (N = 526, 48% male) of college students. The relations between substance use and motives were assessed at each time point, allowing for the examination of these inter-relations over time. Results indicated different trends based on the type of substance. For alcohol use, cross-lagged trends were found between freshman and sophomore year for coping, social, and conformity motives with cross-lagged relations between enhancement motives and alcohol use across all years. However, outside of enhancement motives, cross-lagged relations were not found between sophomore and junior year. In contrast, cross-lagged effects were found for marijuana use and coping, enhancement, and expansion motives between sophomore and junior year, but not freshman year. These results suggest that people's expectations that drinking or smoking marijuana makes activities more reinforcing and helps them cope with distress may perpetuate use. In turn, use itself may enhance these expectations over time. Results have direct implications for treatment, with recommended focus on motives, behavior activation, and healthy coping skills in order to interrupt the cycle of substance use.