Treatment of the premenstrual syndrome is complicated by many factors, but principally by its unknown aetiology. In addition, diagnosis, definition and symptom evaluation methods are unclear. The multitudinous studies of treatment regimens have been largely inconclusive; this is partly due to difficulties in numerically evaluating the symptoms, but more to the lack of appreciation of the marked placebo effect which has been estimated as being up to 50%. The majority of studies have been open studies, and therefore interpretation of the results have been almost impossible. The end result is that great claims have been made for a large number of therapeutic agents on ill-founded evidence. Some of the drugs used in the treatment of the premenstrual syndrome, however, have been better evaluated than others, although even with those studied more extensively results have often been variable. Thus, hormonal agents such as progestagens and oral contraceptives, diuretics, pyridoxine, bromocriptine and danazol have been effective in some studies but not universally so. The latter 2 agents seem to be effective in relieving breast symptoms, but have only a limited effect on other symptoms. It is therefore important to realise that one drug does not cure all patients or all symptoms, although it is often claimed that this is the case. It has been suggested that more than 40% of women suffer from premenstrual syndrome. Greater awareness of the problem, both by patients and doctors, necessitates a more rational approach to therapy.