Background: Despite the availability of many effective treatments, patients with major depression remain at risk for relapse following remission of a depressive episode. The aims of this report are to estimate the relapse rates associated with the acute treatment strategies employed in this study and to investigate demographic and clinical predictors of relapse. Methods: The study sample includes 225 patients who entered the 6-month continuation treatment phase after remitting from an acute depressive episode. Treatment during the acute phase was interpersonal psychotherapy, SSRI (escitalopram), or the combination of the two when monotherapy did not lead to response. Relapse was defined by a Hamilton Depression Rating Scale score ≥15, confirmed by the diagnosis of major depression. The probability of relapsing was modeled using logistic regression. Three separate models were fit with subgroups of covariates. Results: Of the 225 patients, 29 (12.9%) relapsed and 28 (12.4%) discontinued the protocol prematurely. The proportion of patients who relapsed among the group requiring combination treatment to achieve remission was three times as high as among patients who had remitted with monotherapy. In the final logistic regression model, older age, higher baseline HDRS scores, last month (residual) depressive mood spectrum factor score, and requiring combination treatment to achieve remission were each associated with an increased likelihood of relapse. Conclusions: Our results suggest that greater initial depression severity, greater difficulty in stabilizing symptoms, and presence of residual mood spectrum symptoms once remission is achieved are predictive of relapse. Risk of relapse is more likely as age increases, partly because aging confers lower resilience. Depression and Anxiety, 2011. © 2011 Wiley Periodicals, Inc.