A variety of hypotheses have been proposed to explain the premenstrual syndromes. These hypotheses serve as rationales for an equally diverse range of proposed treatments. To investigate these hypotheses, we obtained multiple blood samples across the menstrual cycle in women with well-characterized menstrually related mood disorder and in control subjects. No diagnosis-related differences were observed in the levels or patterns of secretion of progesterone, estradiol, follicle-stimulating hormone, luteinizing hormone, testosterone-estradiol-binding globulin, dehydroepiandrosterone sulfate, dihydrotestosterone, prolactin, or cortisol. Our data suggest that premenstrual syndrome does not represent a simple hormonal deficiency and that the cited rationales for several of the proposed treatments are of questionable merit.