Melatonin was fed daily at 14.00 h to eight non-lactating and six lactating hinds. Feeding was begun on 27 July 1983 (day 1) and continued until 21 September (day 57). Six non-lactating hinds were used as controls. Blood samples were taken about every 10 days from day 42 until day 73. Plasma progesterone and prolactin were determined by radioimmunoassay. Progesterone values of greater than 0·63 μg/1 were taken to indicate the presence of a corpus luteum and that the animal had ovulated. In the melatonintreated, non-lactating group seven of the eight hinds showed ovarian activity compared with two of the six controls (P < 0·01). None of the lactating hinds treated with melatonin showed evidence of ovarian activity. In both the lactating and non-lactating hinds treated with melatonin, prolactin levels were never greater than the detection limit of the assay (7·4 μg/l), while, in the control hinds mean prolactin levels were elevated on all but the last sampling day. It is, therefore, possible to induce early ovarian activity in non-lactating red deer hinds by feeding melatonin. However, although melatonin treatment similarly depressed plasma prolactin levels in both lactating and non-lactating hinds, it was unable to overcome the inhibitory effect of lactation on reproduction.