Summary. In two experiments, mature Merino ewes were fitted with subcutaneous implants containing oestradiol-17\g=b\ or empty implants (control). Peripheral concentrations of LH and FSH were measured during luteal and follicular phases, which were synchronized by means of intravaginal progestagen-impregnated sponges. The ovulation rates of the ewes were determined by laparoscopy, 1 week after luteolysis. In Exp. 1, small implants (3 mm) decreased the LH pulse frequency before luteolysis. In the ewes with oestradiol implants, the amplitude of the preovulatory surge of LH was reduced and the onsets of the LH and FSH surges were delayed. There was no effect of oestradiol on LH baseline, LH pulse amplitude or in the concentration of FSH during the luteal or follicular phases. In Exp. 2, larger implants (10 mm) were inserted for the luteal, follicular, or luteal + follicular phases. The control ewes had empty implants. In the ewes with oestradiol implants, LH pulse frequencies were decreased during the luteal phase. This decrease persisted throughout the follicular phase, even though the oestradiol implants had been removed. There were no differences in the other measures of LH or FSH in the luteal or follicular phases. Ovulation rates were not affected by oestradiol treatment in either experiment, indicating that the regulation of LH pulse frequency is not a critical factor in the determination of ovulation rate in ewes. This was supported by retrospective analysis of the gonadotrophin profiles of single and twin-ovulating ewes, in which all measures of LH were similar between the two groups. However, twin-bearing ewes had slightly higher circulating concentrations of FSH during the luteal phase. As the difference in FSH was not large (13%), and it disappeared after luteolysis, it is difficult to postulate that FSH plays a major role in the final selection of the preovulatory follicle(s) in the ewe.