In 1986 it was reported that a high percentage of women with premenstrual syndrome (PMS) were found to have thyroid hypofunction (TH), mostly subclinical hypothyroidism, as defined by an augmented response of TSH to TRH, and that all affected women had complete relief of PMS symptoms with l-T4 therapy. We studied baseline thyroid function (T4, T3 uptake, T3, TSH, and TSH response to TRH) in 15 normal women (group 1) and 44 women with PMS and treated 22 of the PMS women with l-T4 (group 2; 1.6 μg l-T4/kg dose) and the other half with placebo (group 3) for 2 months in a double blinded protocol. We found no evidence of thyroid dysfunction in group 2 or 3, except for 1 subject with slightly elevated TSH (6.2 μlU/mL) and moderate augmented response to TRH (change in TSH, 65 μIU/mL). During the treatment phase we found a complete relief of symptoms in 6 (27%), a partial relief of symptoms in 6 (27%), and some relief of symptoms in 12 (54%) in group 2. Whereas in group 3, 10 (45%) had complete relief, 5 (23%) had partial relief, and 15 (68%) had some relief of symptoms. These results show that 1) there is no significant thyroid disease in PMS; and 2) l-T4 is no better than placebo in treatment of PMS. We conclude that the high incidence of thyroid hypofunction previously reported in PMS is due to an unusually low TSH level for the limit of the normal range for the TRH stimulation test. © 1990 by The Endocrine Society.