Female tammar wallabies were treated with the dopamine agonist bromocriptine at the end of pregnancy to suppress the peripartum pulse of plasma prolactin. The animals were subsequently observed, and a series of blood samples taken to define the hormonal profiles before and immediately after parturition. Birth was observed in 4/5 control animals and occurred in 8/9 bromocriptine-treated animals. The peripartum peak in plasma PGFM concentrations was not affected by bromocriptine although the pulse of prolactin normally seen at parturition was completely abolished. The timing of luteolysis was apparently unaffected, as plasma progesterone concentrations fell similarly in both treated and control animals immediately after parturition. However, all of the neonates of the bromocriptine-treated animals died within 24 h, possibly because of a failure to establish lactation. Subsequent onset of post-partum oestrus was delayed or absent both in control and in bromocriptine-treated animals, suggesting that the frequent blood sampling and disturbances in the peripartum period interfered with these endocrine processes. It is concluded that both prolactin and prostaglandin can induce luteolysis in the pregnant wallaby, but that the normal sequence of events results from a signal of fetal origin inducing a prostaglandin release from the uterus, which in turn releases a pulse of prolactin that induces a progesterone decline.