The association between neonatal bilirubin exposure and psychoeducational outcome was investigated in a group of grade school children 9 to 11 years old who required neonatal intensive care between 1977 and 1980. Seventy-four children were evaluated with four measures of psychoeducational outcome, including the Kaufman Mental Processing and Achievement Scales, the Beery Visual Motor Integration Test, and the Vineland Adaptive Behavior Scale. A measure of bilirubin binding calculated directly from the albumin concentration correlated significantly with the Kaufman Mental Processing Composite, although other more direct measures of bilirubin exposure (such as maximum serum bilirubin, direct measures of binding, and cumulative bilirubin exposure) did not. Thus, it is possible that the impact on psychoeducational outcome is the result of some other effect of low serum albumin itself, in addition to its ability to bind bilirubin. The correlation of the calculated albumin-determined binding value with the Kaufman Mental Processing Composite suggests that this level, rather than total serum bilirubin, may be more appropriate in determining clinical management.