Objective: Evaluation of thyroid hormone response to a single administration of triiodothyronine (T 3 ) early postnatally to premature infants of ,30 weeks gestational age. Design: A prospective clinical trial with historical control. Methods: Ten infants born ,28 weeks gestational age and ten infants born between 28 and 30 weeks gestational age were given 0.5 mg/kg T 3 intravenously at 12 h after birth. The infants ,28 weeks gestational age were also treated with thyroxine (T 4 ; 8 mg/kg, once daily) during the first 6 weeks of life. Premature infants from a previous trial served as a matched historical control group. Analysis of variance for repeated measurements was performed. Results: For the infants 28±30 weeks gestational age mean plasma T 3 concentrations were significantly higher in the T 3 -treated group P 0X027 for at least 2 weeks, whereas mean plasma levels of T 4 , free T 4 and TSH were comparable. For the infants ,28 weeks gestational age plasma T 3 levels were also significantly different after correction for gestational age P 0X0002Y with either comparable or higher values in the T 3 -treated infants up to 56 days after injection of T 3 . Mean plasma free T 4 levels were lower during the first 3 days and higher or comparable thereafter P 0X0014Y and TSH suppression was more evident in the T 3 -treated infants P 0X003X Conclusion: A single administration of T 3 to premature infants ,30 weeks gestational age early postnatally results in a sustained increase of plasma T 3 levels during the first weeks of life. In infants of 28±30 weeks gestational age this occurs without change in plasma free T 4 levels, whereas in infants ,28 weeks gestational age a transient decrease of plasma free T 4 was present. The increase in plasma T 3 is possibly caused by a T 3 -induced increase of type I deiodinase activity.