As recently claimed, TSH-suppressive therapy with L-T4 may have adverse effects on the heart, but these results have not been consistently confirmed. We assessed cardiac function by clinical, echocardiographic, and ergometabolic criteria in 19 patients (16 women and 3 men) receiving long term L-T4 at a fixed daily dose ranging from 1.8-4.0 microg/kg. The results showed significant alterations in several cardiac parameters suggestive of subclinical hyperthyroidism. In particular, intraventricular septum thickness (10.0+/-1.4 vs. 8.1+/-1.1 mm), left ventricular posterior wall thickness (9.4 1.5 vs. 8.1+/-1.1 mm), end-diastolic dimension (47+/-4 vs. 44+/-3 mm), and left ventricular mass index (102+/-15 vs. 75+/-15 g/m2) were significantly increased compared to values in age- and sex-matched euthyroid controls. Exercise tolerance (expressed as maximal tolerated workload; 102+/-14 vs. 117+/-12 watts), maximal VO2 achieved at peak exercise (maximum VO2, 17.3+/-3.3 vs. 21.9+/-2.5 mL/min x kg), and anaerobic threshold (expressed as a percentage of VO2max, 46.5+/-8.4 vs. 56.2+/-6.6) were significantly reduced in L-T4-treated patients. The L-T4 dose was then reduced to the minimal amount able to keep the serum TSH concentration at 0.1 mU/L or less in 7 patients who were reevaluated 6 months after the initial study. This individual tailoring of the TSH-suppressive L-T4 dose was in all cases associated with normalization of all echocardiographic and ergometabolic parameters. In conclusion, our findings show that abnormalities of heart morphology associated with impaired exercise performance occur as a consequence of long term therapy with fixed TSH-suppressive doses of L-T4, but that these abnormalities improve or disappear after careful tailoring of TSH-suppressive therapy.