We evaluated learning and memory in 50 depressed patients prior to and following 4 week treatment with imipramine compared to 21 normal controls tested at corresponding times. At baseline, the depressives did worse than normals on most memory tasks with the difficult memory tasks, regardless of store, modality or type of task best distinguishing between depressive and normal memory. Following imipramine treatment, responders performed better than nonresponders on the difficult memory tasks, and not significantly differently from controls on most tasks. This, as well as the fact that the responders improved to a greater degree than controls on most measures (in a few cases the difference was statistically significant) and the fact that at 4 weeks complete responders to imipramine did significantly better than partial responders to imipramine, indicates that relief from depression is highly related to improved memory functioning. The finding that complete responders to imipramine were not significantly worse than normal controls suggests that imipramine did not have significant adverse effects on memory.