A recently described in vitro bioassay method for the measurement of biologically active LH [luteinizing hormone] (Van Damme et al. 1974) was applied to the plasma of normally menstruating and post-menopausal women. The specificity of the procedure was established according to the following evidence: parallelism was observed between dose response curves obtained with serial dilutions of a standard LH preparation (HMG 2nd IRP [human menopausal gonadotropin 2nd International Reference Preparation]) and plasma pools collected during the follicular phase, at the LH-peak, during the luteal phase and after menopause; there was no evidence for the presence of any synergistic or antagonistic factor in the various plasma specimens. The assay design used to establish this consisted of assaying the standard and plasma pool separately and then together as a mixture followed by an assessment of the difference (if any) in the potencies obtained. Strict additivity should yield a relative potency of 1.0. Plasma pools which were obtained every 2-3 days throughout the menstrual cycle were assayed using this design against the standard (HMG 2nd IRP) and against a mid-cycle plasma pool obtained at the time of the LH-peak. The latter was also assayed against partially purified plasma fractions obtained from a post-menopausal plasma pool after gel filtration and isoelectric focusing. With the exception of 3 assays, in which the estimates of relative potency were 0.91, 0.94 and 0.95, respectively, in 19 assays of additivity, the fiducial limits always included unity. Non-detectable LH levels were found in the plasma or serum of either hypophysectomized or hypopituitary hypogonadal men or women treated with estrogen/progestogen combined pills. The presence of calf or human serum in the assay medium is an essential requirement for a valid comparison of standard and unknown preparations. In their absence non-parallel dose response curves between plasma and standard were obtained. Other established criteria of reliability (sensitivity and precision) were examined. The method is sufficiently sensitive (3.5-8.0 mIU/ml plasma; HMG (2nd IRP) as standard) for the measurement of LH throughout the cycle. The mean index of precision (.lambda.) in 230 multiple assays was 0.040. The modified bioassay yields valid and reliable estimates of LH when applied to human plasma obtained throughout the menstrual cycle and after menopause.