PURPOSE: To prospectively assess (1) the associations of Effort-Reward Imbalance (ERI), its individual components, and over-commitment with (a) the onset of a Major Depressive Episode (MDE) during a 3.6-year follow-up in a population-based cohort in participants with no current Major Depressive Disorder (MDD) in the beginning of the follow-up (n = 959), (b) incidence of MDD in the subsample of participants exempt from lifetime MDD (n = 490), and (c) the onset of a new MDE (i.e. recurrence) in the subsample of participants with remitted but no current MDD (n = 485), and (2) potential effect modification of burnout on these associations. METHODS: DSM-IV Axis-I disorders were elicited using the semi-structured Diagnostic Instrument for Genetic Studies at each investigation. The ERI Questionnaire was used to measure ERI and overcommitment. Burnout was measured with the Maslach Burnout Inventory General Survey. Serially adjusted logistic regression models were used. The effect of burnout dimensions on these associations was assessed by testing interactions between the ERI and burnout dimensions. RESULTS: (1) ERI was prospectively associated with the onset of MDE, even after adjustment for burnout [OR (95CI) = 1.22 (1.003-1.49)]. (2) The association between ERI and MDD incidence became non-significant after adjusting for burnout. (3) ERI was not associated with recurrence of pre-existing MDD. (4) burnout did not interact with ERI. CONCLUSIONS: Our results support a longitudinal association between ERI and the risk of onset of MDE in the community. Burnout did not modify this effect, but it may partially account for the association between ERI and MDD incidence.