To avoid winter scarcity of fresh goat milk, simple methods of advancing the season of kidding would be commercially valuable. A combination of long-day light treatment followed by melatonin is successful but other aspects of seasonality including coat growth are also reset. To investigate whether effects on breeding season and coat growth can be dissociated, British Saanen dairy goats (no. = 30) were randomly allocated to one of six groups. Control goats were untreated. Goats in the other five groups received 2 months of treatment with 20 h light: 4 h darkness (20L: 4D) and were then given 3 mg melatonin orally at 16.00 hfor 2 months. The treatments started on 11 January (group 1), 29 March (group 2), 14 June (group 3), 30 August (group 4) and 8 November (group 5). Weekly blood samples were taken for the measurement of progesterone to monitor ovarian activity and prolactin. Side patches (10 cm ×10 cm) were clipped monthly from alternate sides and weighed to measure coat growth. The onset of ovarian activity was advanced from a median date of 15 November in controls to 24 May in group 1 (P < 0·01), 16 August in group 2 (P < 0·05) and 18 October in group 3 (? < 0·05). Groups 4 and 5 showed no significant change in the onset of oestrus (median onset dates 8 November and 1 November, respectively). Group 1 goats came into season again with controls in November. In all except group 5, treatment interacted with time to affect significantly the growth of the coat. The pattern of coat growth was most altered (P < 0·001) in groups 1 and 2. This was associated with effects on plasma prolactin concentrations of light stimulation and melatonin suppression. Group 1 goats resynchronized with control goats to show a coat of normal weight in the winter following treatment. Division of milking goat herds into spring and autumn kidding groups is, therefore, a practical possibility, but effects on overall milk yield require study.