Identifying predictors of depression recurrence may highlight targets for relapse prevention. This study tested the hypothesis that strength of implicit and explicit self-depressed associations (SDA) following recovery would lower the threshold for the recurrence of depression. Two main analyses were conducted to test: (i) predictive validity of SDA for recurrence in individuals with a history of depression (at least six months depression free; n = 616, six-year follow-up); (ii) predictive validity of both post-depression SDA and the extent of change in SDA following recovery for recurrence in individuals who recovered in the first two years of the study (n = 220, four-year follow-up). Further exploratory analysis tested the scar model of SDA in participants without a history of depression at baseline. Recurrence rate was 49% in both the first and second analysis. In the first analysis, implicit SDA and explicit SDA were related to recurrence in the unadjusted models, but not when controlling for baseline symptoms. In the second analysis, post-depression explicit SDA predicted recurrence over and above baseline residual depressive symptomatology. There was no support that implicit SDA following depression related to recurrence. However, there was support that stronger explicit SDA following a recent depression increased risk for recurrence.