The use of measuring urinary steroid conjugates in nontimed, randomly collected morning urine samples expressed as a function of creatinine concentration was assessed to monitor ovarian response to pulsatile administration of gonadotropin-releasing hormone in ambulatory patients. This method of evaluating ovarian steroid production provided a convenient, inexpensive, and noninvasive means of monitoring responses to gonadotropin-releasing hormone treatments and documents that clomiphene-resistant amenorrheic patients can be induced to ovulate with appropriate gonadotropin-releasing hormone therapy. Different ovarian responses in the same woman to similar doses and frequencies of gonadotropin-releasing hormone indicate that discrete adjustments of individual doses may be required to facilitate consistent ovulatory responses. The strategy presented here allows for subsequent gonadotropin-releasing hormone therapy in the individual patient to be determined by an objective and quantifiable ovarian response to an initial treatment.