BACKGROUND: Residual depressive symptoms are generally documented as a risk factor for recurrence. In the absence of a specific instrument for the assessment of residual symptoms, a new 25-item Depression Residual Symptom Scale (DRSS) was elaborated and tested for recurrence prediction over a 1-year follow-up. SAMPLING AND METHODS: Fifty-nine patients in remission after a major depressive episode (MDE) were recruited in two centres. They were assessed with the DRSS and the Montgomery-Asberg Depression Rating Scale (MADRS) at inclusion and followed for 1 year according to a seminaturalistic design. The DRSS included specific depressive symptoms and subjective symptoms of vulnerability, lack of return to usual self and premorbid level of functioning. RESULTS: Severity of residual symptoms was not significantly associated with increased risk of recurrence. However, DRSS score was significantly higher among patients with three or more episodes than one to two episodes. Number of previous episodes and treatment interruption were not identified as significant predictors of recurrence. CONCLUSION: The proposed instrument is not predictive of depressive recurrence, but is sensitive to increased perception of vulnerability associated with consecutive episodes. Limitations include small sample size, seminaturalistic design (no standardisation of treatment) and content of the instrument.