Using a randomized double-blind cross-over design, the pharmaco-dynamic and pharmaco-kinetic properties of ‘pure’ follicle-stimulating hormone (FSH) (Metrodin) and human menopausal gonadotrophin (HMG) (Pergonal) were studied in 24 women with polycystic ovary-like disease (PCOD) during induction of ovulation. Fifty-six cycles were stimulated with FSH and 60 cycles with HMG, according to a standard protocol. Gonadotrophins were administered i.v. in a pulsatile fashion using pulse frequencies of either 30 or 120 min. The cycles stimulated with either 30 or 120 min pulse intervals showed no differences among themselves. During the stimulation phase, the FSH and HMG stimulated cycles showed equal and dose dependent FSH concentrations (mean ± SD). The luteinizing hormone (LH) concentrations (mean ± SD) were also equal but unchanged compared to the mean basal concentration. The LH, FSH, total urinary oestrogen excretion, and testosterone profiles (mean ± SD) obtained from cycle days −10 to 0 as well as the pregnanediol profiles obtained from cycle days 0 to +14 showed no differences either. The occurrence of an endogenous preovulatory LH surge was significantly more frequent in the cycles stimulated with a pulse interval of 30 min compared to the cycles stimulated with a pulse interval of 120 min. The addition of LH as provided in HMG did not influence the FSH threshold concentration above which initiation of follicular growth occurred, since no differences were found in the FSH ‘stable’ concentrations between FSH and HMG stimulated cycles. However, intra- and inter-individual variation in the FSH ‘stable’ concentration at which follicular growth was initiated became obvious. It has been hypothesized that either diminished circulating bioactive FSH or intrafollicular paracrine factors may influence the FSH threshold concentration above which the ovary responds with follicular growth.