In order to evaluate the relationships between endogenous opioid activity and premenstrual complaints, we subjected three groups of patients in the mid (days 8-12 prior to menses) and late (days 1-5 prior to menses) luteal phases of the cycle to a naloxone test and some of the patients to a luteinizing-hormone-releasing hormone (LHRH) test. The premenstrual syndrome (PMS) group was composed of nine patients complaining of dizziness, irritability and depression close to menses for at least three years. The menstrually related migraine (MM) group was composed of 15 patients complaining of premenstrually related migraine. The common migraine (CM) group was made up of 16 women suffering from common migraine for years whose attacks occurred independently of menstrual cycle events. A group of seven fertile women served as controls. Every two days the patients filled out the Menstrual Distress Questionnaire for evaluation of their complaints. After the evaluation of spontaneous LH pulsatility for one hour, 4 mg of naloxone was injected as a bolus, and samples were collected every 15 minutes for 2 hours. Both estradiol (E2) and progesterone (P) were measured in basal samples from each naloxone test. LH responsiveness to LHRH was similar in the mid and late luteal phases and did not change between groups. In the mid luteal phase the LH response to naloxone in PMS and MM patients was similar to that in normal subjects, while CM patients had impaired LH secretion. In the premenstrual phase only the controls maintained an LH responsiveness similar to that observed in the mid luteal phase, while both PMS and MM lost the naloxone-induced LH release.(ABSTRACT TRUNCATED AT 250 WORDS)