We evaluated the predictive value of the thyrotropin (TSH) response to thyrotropin-releasing hormone (TRH) in 32 depressed outpatients completing a double-blind placebo-controlled trial of s-adenosyl-l-methionine (SAMe), which failed to show any significant difference between SAMe and placebo. Treatment response was defined as the change in Hamilton Rating Scale for Depression (HRSD-24) score between baseline and the end of the six-week trial. Subjects with TSH response outside the normal range (7-25 uU/ml) had a significantly greater response than patients with a normal response. There was also a significant correlation between absolute deviations from the mean TSH response (16 uU/ml) and changes in HRSD-24 scores.