ABSTRACT. To assess the influence of treatment on the development of the somesthetic pathway in infants with congenital hypothyroidism receiving early treatment, median nerve somatosensory evoked potentials were measured during the 1st y of life.Twenty-nine infants were studied with six to seven somatosensory evoked potential tests per infant. The cervical latency (N13) divided by arm length and the first (N19) and second (N32) cephalic latencies as well as N13-N32 latency were measured. At diagnosis, all components showed a small but significant delay, which was not related to thyroxine (T4) levels before treatment. During treatment, T4 ranged from 50 to 290 nmol/L. At 12 mo, the cervical latency divided by arm length had normalized, whereas N19 and N13-N32 were more abnormal than at diagnosis. For N19, these abnormalities were related to a slow initial rise of T4 (1100 nmol/L after 1 wk of treatment) and the initial N19 values. Abnormal N13-N32 values were associated with high T4 values during treatment (>200 nmol/L) and the type of congenital hypothyroidism (partial or total deficiency in T4 production). Induction of therapy with I-triiodothyronine rather than I-thyroxine and the occurrence of low T4 values ( e l 0 0 nmol/L) after the 4th wk of therapy had no such effect. Our data suggest that, for normal CNS development, euthyroidism should be reached as soon as possible by adequate induction therapy. Thereafter, T4 supplementation should be strictly dosed, keeping the serum T4 values within narrow limits around the mean normal for age, because overtreatment, like initial undertreatment, may lead to CNS abnormalities at the end of the first year. (Pediatr Res 34: 73-78,1993) Abbreviations CHT, congenital hypothyroidism I-T4, levo-thyroxine 1-T3, levo-triiodothyronine I-T4/f, fast initial serum T4 rise on I-T4 induction I-T4/s, slow initial serum T4 rise on I-T4 induction PD, partial deficiency in T4 production TD, total deficiency in T4 production PMA, postmenstrual age SEP, somatosensory evoked potentials N13/AL, latency to first negative peak in cervical lead divided by arm length N19, latency to first negativity in cephalic lead N32, latency to second negativity in cephalic lead N60, latency to third negativity in cephalic lead P22, latency to first positivity after N19 T4>200, group with serum T4 >200 nmol/L on all occasions T 4~2 0 0 , group with serum T4 ~2 0 0 nmol/L on all occasions bl, slope for period T4 ~2 0 0 nmol/L b2, slope for period T4 >200 nmol/L The major objective of the national screening programs for CHT is the prevention of neurologic damage. On the whole, the results of the screening seem gratifying, but there are still unsatisfying aspects. Several studies (1-3) report CHT infants displaying minor neurologic dysfunctions and subnormal IQ at later age despite early diagnosis and treatment. Other studies report normal findings (4-6).Generally, the neurologic deficits are ascribed to perinatal thyroid hormone deficiency, because they were found to be related to very low initial T4 val...