Conflicting results have recently been published about the benefits of combined T(4) and T(3) in treating hypothyroid patients. However, these studies may have been underpowered to detect differences in psychological well-being specifically related to T(4) replacement. We conducted a large, double-blind, randomized controlled trial of partial substitution of 50 microg T(4) by 10 microg T(3) vs. the original dose of T(4) in 697 hypothyroid patients. Thyroid function showed a rise in TSH (132%), a fall in free T(4) (35%, P < 0.001), and unchanged basal free T(3) levels (P = 0.92). At 3 months, there was a large (39%) placebo effect improvement in psychiatric caseness defined by the General Health Questionnaire (GHQ) 12 score in the control group compared with baseline, and this was sustained at 12 months. Differences vs. the intervention (T(3)) group were more modest with improvements in GHQ caseness (odds ratio, 0.61; 95% confidence interval, 0.42, 0.90; P = 0.01) and Hospital Anxiety and Depression questionnaire-anxiety scores at 3 months (P < 0.03) but not GHQ Likert scores, Hospital Anxiety and Depression questionnaire-depression, thyroid symptoms, or visual analog scales of mood and the initial differences were lost at 12 months. These results may be consistent with a subgroup of patients showing transient improvement after partial substitution with T(3) but do not provide conclusive evidence of specific benefit from partial substitution of T(4) by T(3) in patients on T(4) replacement. They also emphasize the large and sustained placebo effect that can follow changes in thyroid hormone administration.