Depression is the leading cause of global disability with over 300 million people suffering world-wide. Estimates suggest that up to two thirds of patients do not recover following their first antidepressant treatment and up to one third do not recover after multiple treatments. Therefore, it is critically important to identify factors that predict recovery and reduce risk of relapse. Current methods in genetic epidemiology focus on predictors of mental illness onset. While this is crucial to prevent new diagnoses, it does little to help individuals already suffering. Therefore, the Recover project aims to extend current genetic epidemiology methods to better understand recovery from depression. The methods developed here will begin with a focus on depression but can also be extended to other mental illnesses. First, we will develop trajectories of depression using continuous longitudinal measures in two critical time points, 1) adolescence and early adulthood and 2) during and post pregnancy. Second, using these trajectories as outcomes we will explore many modifiable predictors of recovery. Third, we will use cutting-edge causal inference techniques to test whether or not these predictors are causal. And finally, we will develop novel technologies to capture fine-grained fluctuations in mood. Taken together, this work will lead to better interventions and inform adjuncts to treatment having real impact for the growing number of individuals suffering from depression.