The effect of recombinant bovine FSH (rbFSH) administration on serum FSH and follicular growth was determined in 16 GnRH-immunized heifers (primary day 0 and booster day 28 using 0·4 mg human serum albumin (HSA) – Cys-Gly-GnRH in DEAE-dextran). All heifers produced GnRH antibody titres (mean 58·8 (s.e. 2·8)% binding at 1: 640 serum dilution on day 42), follicular growth ceased (follicles ≤ 4 mm) between days 33 and 49, progesterone concentrations were ≤ 0·5 ng/ml in all heifers by day 53, and pulsatile LH secretion was markedly reduced (v. luteal phase control heifers) on day 63. Heifers were blocked by antibody titre and LH concentrations on day 63 and assigned (no. = five or six per treatment) on day 78 to: (1) control (saline given i.m. four times per day for 4 days), (2) 0·5 mg equivalent (USDA bFSH BP 1) of rbFSH (Granada Biosciences, TX) given i.m. four times per day for 6 days (12 mg in total) or 3) 1·5 mg rbFSH given i.m. four times per day for 4 days (24 mg in total). Ovaries of heifers were examined by ultrasound and blood samples were collected at 3- to 6-h intervals from days –1 to 10 and then twice a day until follicles ≥ 5 mm were no longer detectable. The increase in serum FSH concentrations was different (P < 0·05) between treatments (mean peak FSH above pre-treatment baseline concentrations at 102 h after first rbFSH were 1·4 (s.e. 0·8), 14·5 (s.e. 3·3) and 33·6 (s.e. 5·3) ng/ml for treatments 1 to 3, respectively). The rate of decline in FSH from 102 to 302 h was not different between treatments 2 and 3. There was no detectable follicular growth in heifers given saline injections. Following either rbFSH treatment, medium (MF; 5 to 9 mm) and large (LF; ≥ 10 mm) follicles grew. For heifers in treatments 2 and 3, there was no difference in the interval from start of rbFSH until the first appearance of MF (4·2 (s.e. 0·7) v. 3·3 (s.e. 0·3) days) or LF (5·3 (s.e. 0·3) v. 4·8 (s.e. 0·8) days) or in the duration of the presence of MF and/or LF (13·2 (s.e. 4·5) v. 16·6 (s.e. 2·6) days). Increasing dose of rbFSH increased (P < 0·05) the mean maximum number of MF (4·3 (s.e. 1·1) and 9·2 (s.e.0·9) for treatments 2 and 3) and LF (2·0 (s.e. 1·1) and 8·4 (s.e. 1·2)). In summary, low doses of rbFSH induced a small cohort of follicles to grow and develop without selection or development of a single dominant follicle, whereas high doses of rbFSH resulted in large numbers of both MF and LF.