The involvement of the uterus in early regression of corpora lutea induced by the ram effect was studied in seasonally anoestrous Barbarine ewes. In experiment 1, group I was only submitted to the male effect (control, n = 15) while group II (n = 14) was injected every 12 h with flunixin meglumine, a PGF2 alpha synthetase inhibitor (finadyne), from day 3 to day 6 (day 0: day of ram introduction). The preovulatory LH surge appeared at the same time (around 21 h after ram introduction) in both groups. Finadyne treatment significantly decreased 13-14-dihydro-15-keto-PGF2 alpha (PGFM) pulses (1.3 +/- 0.3 versus 0.4 +/- 0.2; P < 0.05), the number of short cycles (50 versus 14%, P < 0.05), and provided a single peak of oestrus, 15-16 days after the introduction of the rams, instead of between day 14 and day 23 (P < 0.01). In experiment 2, 17 hysterectomized ewes were allocated into two groups: group III (n = 8) was injected with oil and group IV (n = 9) received an intramuscular injection of 20 mg of progesterone immediately before introduction of rams. An additional group of intact ewes was used as control (group V, n = 9). Hysterectomy did not affect the ovulation response to the ram effect, but completely suppressed short cycles (0 versus 78%, P < 0.001). The preovulatory LH surge was delayed in hysterectomized females (36.0 +/- 14.1 versus 16.6 +/- 11.4 h; P < 0.004). Treatment with progesterone significantly (P < 0.01) increased the interval between introduction of rams and the preovulatory LH surge. In conclusion, suppression of short cycles by hysterectomy and an inhibitor of PGF2 alpha synthetase suggest that the uterus is essential for determining the lifespan of ram-induced corpora lutea and that premature release of PGF2 alpha is the cause of early luteal regression. The hypothesis that lower secretion of progesterone before D5 could be the initial cause of the premature induction of the luteolytic signal is discussed.