Two experiments involving 24 and 54 Australian Merino ewes were conducted in which the establishment of a cervical population of spermatozoa and several endocrinological events were studied after several regimens for the synchronization of oestrus. Intravaginal sponges impregnated with 500 mg (Exp. 1) or 200, 400 or 600 mg (Exp. 2) progesterone resulted in the maintenance of plasma progesterone concentrations of 1.5-4.9 ng/ml over a 12-day insertion period compared with 1.9-6.9 ng/ml during dioestrus in control ewes. In Exp. 1 basal concentrations of less than or equal to 0.25 ng/ml plasma were attained by 4 h after sponge withdrawal and this decline was much more rapid than in normal luteolysis. This was associated with fewer spermatozoa recovered from the cervix 2 h after insemination, and PMSG had no significant effect. In Exp. 2 injection of a supplementary dose of progesterone at sponge withdrawal resulted in a rapid increase in plasma progesterone concentrations followed by an equally rapid decrease and an attenuation of the rise in plasma oestradiol-17 beta, the LH surge, and the onset of oestrus. The numbers of spermatozoa recovered 4 h after insemination were not increased, and PMSG had no significant effect. Two factors were significant, namely the dose of progesterone in the sponge (600 mg greater than 400 or 200 mg, P less than 0.05) and stage of oestrus when inseminated (mid- or late oestrus greater than early). The data demonstrated that an adequate dose of progesterone/progestagen incorporated into intravaginal sponges and accurate timing of insemination relative to the LH surge are the most important factors involved in penetration of the cervix by spermatozoa.