The molecular basis for variations in resting metabolic rate (RMR) within a species is unknown. One possibility is that variations in RMR occur because of variations in uncoupling protein 2 (UCP-2) and uncoupling protein 3 (UCP-3) expression, resulting in mitochondrial proton leak differences. We tested the hypothesis that UCP-2 and -3 mRNAs positively correlate with RMR and proton leak. We treated thyroidectomized and sham-operated mice with triiodothyronine (T(3)) or vehicle and measured RMR, liver, and skeletal muscle mitochondrial nonphosphorylating respiration and UCP-2 and -3 mRNAs. T(3) stimulated RMR and liver UCP-2 and gastrocnemius UCP-2 and -3 expression. Mitochondrial respiration was not affected by T(3) and did not correlate with UCP-2 and -3 mRNAs. Gastrocnemius UCP-2 and -3 expression did correlate with RMR. We conclude 1) T(3) did not influence intrinsic mitochondrial properties such as membrane structure and composition, and 2) variations in UCP-2 and -3 expression may partly explain variations in RMR. One possible explanation for these data is that T(3) stimulates the leak in vivo but not in vitro because a posttranslational regulator of UCP-2 and -3 is not retained in the mitochondrial fraction.