The amount of prolactin has been determined in serial blood samples taken over 24 h from 20 pre- and 9 postmenopausal women volunteers. All women had a large increase in prolactin at night (24.00 h-03.00 h). A much smaller rise in prolactin occurred in the evening (18.00 h-20.00 h) which was just significantly different (p less than 0.05) from levels found in the afternoon. Prolactin concentrations at any given time in the 29 women were linearly correlated with corresponding levels in either of the adjacent time intervals. This correlation was highly significant (p less than 0.001) throughout 24 h, implying that a woman with a relatively high (or low) blood prolactin in the afternoon would have a similarly high (or low) concentration in the evening or at night. Premenopausal women had higher amounts of prolactin than postmenopausal volunteers throughout the day. Parity was found to be inversely related to prolactin levels in pre- and postmenopausal women; this effect was especially marked in the night peak of prolactin. In postmenopausal women an average of 33 years had elapsed since the birth of the last child, which implies that childbirth has a life-long effect on reducing prolactin levels. Thus the protective effect of early child-bearing and multiparity against developing breast cancer could be at least partly explained by these effects on serum prolactin levels.