The reproductive efficiency of Friesian dairy cows was investigated in a three (oestrous synchronisation technique) x two (seasons of the year) factorial design. The 90 primiparous and multiparous cows (winter, n = 42; summer, n = 48) were allocated at random to three synchronisation treatments (n 30 cows per treatment). In treatment 1 (GPG), the cows were administered 15 mg PGF(2alpha) i.m. at 30 +/- 3 days postpartum, 100 mug GnRH i.m. at 51 +/- 3 days and 15 mg PGF(2alpha) 7 days later. A second 100 mug dose of GnRH was given after, further 2 days and fixed time AI occurred 16-20 h later. In treatment 2 (PG-PG), 15 mg PGF(2alpha) was administered i.m. to each cow on three occasions at successive- 14 days interval starting at 30 3 days postpartum and the cows were inseminated at observed oestrus following the third dose of PGF(2alpha). Cows in treatment 3 (PG) had a single administration of 15 mg PGF(2alpha) i.m. at 57 3 days postpartum and were inseminated as in treatment 2. Mean daily ambient temperature was 10.9degreesC in winter (November-March) and 20.2degreesC in summer (June-October). The cows were confined in an open-fronted shed and had ad libitum access to a complete diet with a 37:63 forage to concentrate ratio. Body condition score was assessed at 57 +/- 3 days postpartum. Cow rectal temperature at insemination, milk yield, reproductive data and climatic variables Were recorded. Blood samples were collected for progesterone assay on days 4, 11, 18, 25, 32, 39 and 46 post-AI from 54 of the cows (19 GPG; 17 PG-PG; 18 PG). Pregnancy rate to first Al was 36.7% (11/30) for GPG and 16.7% (5/30) for both PG-PG and PG treatments. The difference was not significant. The cumulative pregnancy rate after third Al were GPG 83.3% (25/30), PG-PG 60.0% (18/30) and PG 60.0% (18/30; P < 0.057). The cumulative pregnancy rate for cows inseminated in the winter (81.0%; 34/42) was higher (P < 0.01) than for those inseminated in the summer (56.3%; 27/48). The interval from calving to first service was shorter (P < 0.05) in treatment PG-PG (65.4 +/- 1.3 days) than in PG (69.2 +/- 1.3 days). Mean plasma progesterone concentrations post-AI of pregnant cows were higher (P < 0.001) for GPG cows than those for PG-PG and PG cows. Plasma progesterone levels of pregnant cows tended to be higher (P = 0.087) in winter than in summer. In conclusion, although the cumulative pregnancy rate was higher for GPG cows, it may be appropriate to correct the nutrition and management of the herd before resorting to synchronisation techniques to improve animal reproductive performances. (C) 2002 Elsevier Science B.V. All rights reserved.