1. There are major changes in progesterone, oestrogen, cortisol and beta-endorphin level associated with parturition, and as all these can be psychoactive it is likely that they contribute to the mood changes that can occur at this time. However evidence for their involvement is, at present, indirect. 2. Postnatal depression itself appears to be a heterogeneous condition with different times of onset, and it is probable that various biological and social factors play a role to a differing degree in different individuals. 3. About half of postnatal depression appears to arise in the first two weeks after childbirth. Some cases follow a period of early euphoria. 4. A different subgroup is associated with thyroid dysfunction, which peaks two to five months postpartum. 5. The tyramine test does not predict vulnerability to postnatal depression. 6. It is suggested that in future research the time course of onset of the depression, and the nature of the mood changes that occur in the first postpartum week, are investigated as possibly relevant variables.