Summary. The specific requirement for FSH in the final stages of preovulatory follicle development was assessed in seasonally anoestrous ewes given 2-h injections of GnRH (250 ng/injection), with (N = 10) or without (N = 10) concurrent treatment with bovine follicular fluid (bFF: 2 ml given i.v. at 8-h intervals). Treatment with bFF significantly (P < 0\m=.\01)suppressed plasma FSH concentrations, but, at least for the first 30 h of treatment, did not influence the magnitude of GnRH-induced LH episodes (mean max. conc. 3\m=.\00 \ m=+-\ 0\m=.\39and 3\m=.\63 \ m=+-\ 0\m=.\51 ng/ml for bFF-treated and control ewes, respectively). Of 10 animals treated with GnRH for 72 h, 5/5 control ewes showed oestrus and ovulated whereas 0/5 bFF-treated ewes showed oestrus or ovulated in response to GnRH treatment. There was, however, a transient (13\m=.\2\m=+-\ 1\m=.\0h) increase in plasma LH concentrations in the ewes given bFF (mean max. conc. 4\m=.\64\m=+-\1\m=.\57 ng/ml), which was coincident with the preovulatory LH surge recorded in animals given GnRH alone. In 10 GnRH-treated ewes slaughtered after 32 h of treatment, the mean diameter of the largest antral follicle was significantly (P < 0\m=.\001) greater in control ewes (5\m=.\92 \m=+-\0\m=.\17 mm) than in animals that were also given bFF (3\m=.\94\ m=+-\ 0\m=.\14 mm). In addition, the incidence of atresia in the 3 largest antral follicles present at this time was greater in bFF-treated ewes. These results show that, when plasma FSH concentrations are suppressed by administration of bFF, although the magnitude of GnRH-induced LH episodes is unchanged, preovulatory follicular development is impaired and ovulation does not occur. This may be indicative of a specific requirement for FSH in the final stages of preovulatory follicle development, or due to direct inhibitory effects of bovine follicular fluid.