Phototherapy for neonatal hyperbilirubinemia was introduced to our medical center in March 1971. To assess the influence of phototherapy on subsequent cognitive outcome, we compared the intelligence test scores at 17 years of age of subjects born 4 months before and 10 months after the introduction of phototherapy. The intelligence quotient score (mean +/- SE) at 17 years for 84 subjects with severe neonatal hyperbilirubinemia was 108 +/- 2 for those treated by phototherapy and 107 +/- 2 for controls matched for gestational age and birth weight. The confounding effect on intelligence quotient scores of perinatal factors (bilirubin concentrations, gestational age, birth weight, Apgar score) and demographic characteristics (ethnic origin, socioeconomic status, paternal education) was taken into account in a multiple logistic regression analysis by using a General Linear Models procedure. Phototherapy was found to have no independent effect on intelligence quotient scores after adjustment for the effect of confounding factors. We conclude that for full-term newborn infants with neonatal hyperbilirubinemia, phototherapy had neither a beneficial nor an adverse effect on intellectual ability in late adolescence.