The objective of the current investigation was to examine the prevalence of pathological gambling (PG) in a psychiatric sample with a history of mood disorder, and the concurrent and longitudinal association of PG and mood disorder symptoms according to retrospective report. A total of 275 (100 male, 175 female) psychiatric outpatients in Ontario, Canada, with a lifetime diagnosis of a depressive (n = 138) or bipolar disorder (n = 137), completed the Canadian Problem Gambling Index, South Oaks Gambling Screen and Longitudinal Interval Follow-up Evaluation. Correlational and cross-lagged panel analyses evaluated the relation between PG and mood disorder symptom course. The prevalence of PG was elevated within patients with a mood disorder; there was no difference across diagnosis. Concurrent PG and mood disorder symptoms were positively correlated; however, longitudinal analyses revealed no evidence for an association between PG and mood disorder symptoms when symptom stability was taken into account. Despite the elevated prevalence of PG within mood disorders, and the concurrent association between PG and mood disorder symptoms, no direct association was found between these types of pathology. Prospective designs and intervening variables are required to advance understanding of the etiological associations between these disorders.