The effects of long-term treatment with physiological doses of oestradiol or oestradiol plus progesterone on plasma gonadotrophin levels and pituitary content of LH and gonadotrophin-releasing hormone (GnRH) receptors were studied in ovariectomized-hypothalamo-pituitary disconnected ewes given 250 ng pulses of GnRH every 2 h (i.v.). A pilot experiment showed that 3 cm long Silastic implants (s.c.) reduced both LH pulse frequency and pulse amplitude in long-term (>6 months) ovariectomized ewes. The main experiment was conducted over 3 weeks in ovariectomized-hypothalamo-pituitary disconnected ewes that had received pulsatile GnRH replacement for 1 week after pituitary surgery. Group 1 (n = 5) received GnRH pulses alone throughout the study. Group 2 (n = 6) received oestradiol in week 2 and oestradiol plus progesterone in week 3 and in group 3 (n = 6) the steroid treatments were reversed. Oestradiol reduced (P <0.05) the mean (± S.E.M.) amplitude of LH in pulses in group 2 (from 8.2 ± 1.6 to 5.0 ± 0.5 μg/l) and group 3 (from 11.6 ± 1.2 to 9.3 ± 1.0 μg/l); an additional effect of progesterone was seen in group 2 but not group 3. The amplitudes of the LH pulses did not change in the control ewes. Plasma concentrations of FSH were reduced by approximately 50% by the oestradiol treatments with no additional effects of progesterone. There was no effect of steroidal treatment on pituitary content of LH or pituitary levels of GnRH receptors. We conclude that long-term oestradiol treatment, with or without progesterone, reduces plasma amplitudes of LH pulses by a direct pituitary effect, but the magnitude of this effect was less than that observed on GnRH secretion in short-term ovariectomized ewes in an earlier study. The reduction in plasma LH pulse amplitude is not due to a direct pituitary effect of these steroids on GnRH receptor number.