The pattern of episodic gonadotropin release was studied in 15 normal female volunteers during the luteal phase of the menstrual cycle with 24 h of blood sampling for follicle-stimulating hormone (FSH) and luteinizing hormone (LH) levels at 10-min intervals. Six subjects (two in the early, two in the mid-, and two in the late luteal phase) also had each of these specimens processed for progesterone levels. A progressive slowing of LH pulsations was present across the luteal phase with the mean LH pulse frequency declining from 15.2 pulses/24 h in the early to 8.4/24 h in the late luteal phase. A trend towards reduction in the amplitude of LH pulses was also observed (12.3 +/- 2.2 SD mIU/ml in the early vs. 8.6 +/- 3.4 mIU/ml in the late luteal phase; NS). In addition, LH pulses of heterogeneous amplitude were identified during the same 24-h study. The mean +/- SD of the larger and of the smaller LH pulses was 16.9 +/- 4.7 and 2.3 +/- 1.0 mIU/ml, respectively (P less than 0.001). While the slowing of the frequency of all LH pulses correlated well (r = 0.80, P less than 0.001) with the day of the luteal phase and poorly with the actual plasma progesterone levels, the incidence of the small LH pulses was highest in the mid-luteal phase and correlated well with the mean progesterone plasma levels (r = 0.63, P less than 0.01). In the early luteal phase, the pattern of progesterone secretion was stable over the 24-h studies and showed no relationship to episodic LH release. In contrast, in the mid- and late luteal phase, plasma progesterone concentrations rapidly fluctuated during the 24-h studies from levels as low as 2.3 to peaks of 40.1 ng/ml, often within the course of minutes. Progesterone increments closely attended episodes of LH release, as documented by the significant (P less than 0.05) cross-correlation between LH and progesterone levels, at time lags of 25-55 min. The results of this study indicate that in the human luteal phase: (a) the frequency of pulsatile release of LH declines progressively and correlates well with the duration of exposure to progressively and correlates well with the duration of exposure to progesterone; (b) the amplitude of LH pulses varies with the appearance of an increased percentage of smaller pulses correlating well with the acute level of progesterone; (c) in the early luteal phase, the pattern of progesterone secretion is stable; (d) in the mid- and late luteal phase, progesterone secretion is episodic, and correlates with LH pulsatile release; and (e) single progesterone estimations in the mid- and late luteal phase do not accurately reflect corpus luteum adequacy.