Psychotic symptoms are not included under accepted definitions of prem
enstrual syndrome (PMS). We present a 14-year-old girl with PMS, who d
eveloped a late luteal cyclic psychosis during two consecutive premens
trual periods, which resolved completely after the onset of menses. Sh
e was treated with dehydroxyprogesterone for two cycles, and later wit
h placebo for the next three consecutive cycles. Psychotic symptoms di
d not reappear following two psychotic cycles, and the PMS resolved wi
thin the next menstrual cycle. We suggest that cyclic psychoses associ
ated with the menstrual cycle may be a specific benign entity, not inc
luded under the recognised functional psychoses. In some cases these p
sychoses could be classified as a subgroup of PMS.