The plasma prolactin levels of 98 healthy women 22 to 65 years old showed a strong inverse correlation with age. This relationship persisted for older premenopausal and postmenopausal women (aged 30 to 65 years) whose results were compared with those of 34 premastectomy early breast cancer patients, another 38 studied four to six weeks after mastectomy, and 38 with advanced disease (age range, 30 to 77 years). There was no correlation between age and plasma prolactin in any of the breast cancer groups. When patients and controls were classified according to their menopausal status, the premastectomy and postmastectomy early breast cancer groups both showed significantly higher prolactin concentrations than the corresponding controls (P less than 0.001) in all cases). These abnormalities were particularly prominent in the patients studied after surgery, although the difference was significant only for the postmenopausal groups (P less than 0.05). Plasma cortisol and urinary free cortisol levels were normal in postmastectomy patients, suggesting that stress was not the cause of elevated prolactin levels. In advanced breast cancer, elevated plasma prolactin concentrations were found only in the postmenopausal patients.