Many women experience cyclical changes in well-being. In a small proportion cyclicallyoccurring negative changes are serious enough to interfere with daily activities and have bewme known as the premenstrual syndrome (PMS). The belief that the aetiology of premenstrual symptoms is ultimately hormonal, even if mediated through someother mechanism, has been a long-standing one in the literature on the subject. The area of research is reviewed, highlighting the problem of data interpretation. Methodological inconsistencies are common, for instance, variations in the definition and diagnosis of PMS and the endocrinological parameters assessed. This review suggests that the timing of any degree of premenstrual change is related to the ovarian cycle. However, the severity of symptoms experienced, which distinguishes the woman with PMS from one without the syndrome, does not appear to be related to any hormonal parameter so far assessed. Further investigations of the psychology of the menstrual cycle are needed before the aetiology of PMS will be determined.