Incubation is the process by which life events influence the onset of psychiatric disorder after an appreciable delay. It has long been recognised clinically. In this paper we use data from the Camberwell Collaborative Depression Study to see whether incubation effects can be demonstrated in depressive illness. We used a novel adaptation of survival analysis for this purpose. The results suggest that incubation does occur in depressive disorder, that it is much less important than the effect of life events close to onset, that it is apparent in women but not in men, and that it is no more evident preceding endogenous than neurotic symptom patterns.