BACKGROUND: To test the effects of progressively decreasing dosages of exogenous LH we combined various amounts of HMG, containing FSH, LH and HCG, and highly purified (HP) FSH to treat 120 GnRH agonist-suppressed infertile female patients as candidates for controlled ovarian stimulation (COS). METHODS: Subjects were randomly assigned to four treatment groups that received the following daily i.m. gonadotrophin regimens: A, FSH 150 IU only; B, FSH 150 IU and LH activity 37.5 IU; C, FSH 150 IU and LH activity 75 IU; D, FSH 150 IU and LH activity 150 IU. FSH dose adjustments were allowed only after the 14th treatment day. Monitoring included transvaginal ultrasound at 2-day intervals and daily determinations of LH, FSH, estradiol (E2), progesterone, testosterone and HCG. RESULTS: Duration of COS was significantly shortened in patients receiving at least 75 IU daily of LH activity. Small (r = −0.648, P < 0.0001) and serum HCG levels (r = −0.272, P < 0.01) but not to serum LH levels. Serum testosterone levels were positively correlated to the LH activity dose administered (r = 0.313, P < 0.001), while serum progesterone levels were positively correlated to the FSH dose administered (r = 0.447, P < 0.00001) but not to the LH activity dose administered. CONCLUSIONS: Firstly, HCG content considerably contributes to HMG activity; secondly, menotrophin LH activity content can reduce in a dose-dependent manner the occurrence of small pre-ovulatory follicles; and finally, contrary to common belief, enhanced FSH stimulation rather than LH activity appears to cause premature follicle luteinization during COS.