Biological and immunological measurements of human luteinizing hormone (hLH) were performed with plasma samples obtained by continuous withdrawal over a 10-hour period in various phases of the normal menstrual cycle. A pulsatile pattern of hLH discharge was demonstrated by both the in vitro bioassay and radioimmunoassay throughout the menstrual cycle, although the frequency and the amplitude of the pulse differed according to the menstrual stage. The decreased half disappearance time of immunoreactive hLH (83.7 min with 95% confidence limits at 64.3 and 103.1 min), in comparison with that of bioactive hLH (122.9 (98.2: 147.6) min), and the rapid recovery of the depressed biological to immunological activity ratio (B/I ratio) at the peak of the pulse (0.87 (0.77:0.96) in the periovulatory phase indicated the concomitant discharge of hLH subunits with hLH. The differences in the half disappearance times of bioactive hLH in the follicular and luteal phases (103.6 (94.7:112.5) min and 87.2 (72.5: 101.9) min, respectively) and in the B/I ratios at the peaks of the pulses in these two phases (1.17 (1.05:1.30) and 1.03 (0.94:1.12), respectively) suggested the selective secretion of hLH subpopulations with various biological and immunological potencies and with various clearance rates. The significantly depressed B/I ratios at the trough in the luteal phase (0.58 (0.44:0.76) were sufficient to question the overall reliability of the hLH-RIA procedure employed, especially when applied to low hLH levels, even after several methodological improvements.