There is a high comorbidity between symptoms of depression and cannabis and alcohol use in civilian and veteran populations. Prospective studies attempting to clarify the directionality of these comorbidities have yielded mixed results. Further, the relations between these constructs and impulsive personality, particularly negative urgency (NU, the tendency to act rashly when experiencing emotional distress) warrants further attention, as NU relates to symptoms of depression and alcohol and cannabis use. Importantly, NU partially accounts for the association between symptoms of depression and cannabis and alcohol problems in cross-sectional studies. This study examined alternative theories of directionality in order to better understand the longitudinal associations between symptoms of depression, NU, and cannabis or alcohol use. Three semiannual waves of data (baseline, 6-month, and 12-month) were collected in parallel assessments from a sample of Operation Enduring Freedom, Operation Iraqi Freedom, and Operation New Dawn veterans (N = 361). Autoregressive cross-lagged panel models were used to test four alternative theory-driven models about the longitudinal associations between the interaction of symptoms of depression and NU and cannabis or alcohol use. Models revealed unique direction of effects specific to each substance, such that the interaction between symptoms of depression and NU at 6 months postbaseline predicted more alcohol use at 12 months postbaseline, whereas more cannabis use at 6 months postbaseline predicted more severe symptoms of depression at 12 months postbaseline. Results suggests alternate directions of effect for cannabis and alcohol use. Future research should examine these patterns over wider assessment periods in order to see more variability and change over time.