Nineteen early-treated phenylketonuric patients, whose diet was discontinued between 4.5 and 13 years of age, and who have been off the diet for 12-28 years, were reassessed in 1992-93. There was little change in mean IQ between end of diet and follow-up, less than one IQ point on the average, with no change for any individual exceeding 12 IQ points. Both prior and current IQ correlated slightly negatively with mean phenylalanine (Phe) concentration, and positively with parents' education. The phenylalanine level at follow-up was significantly lower on average by about 900 μmol/L. Five of the subjects (26%) have evidence of mental disease. However, the data suggest that the discontinuation of the diet did not cause intellectual deterioration. Nonetheless, the patients' intellect cannot be the only consideration for maintenance of diet. The occurrence of psychopathology among phenylketonuric patients and the possible unknown effects of toxic elevation of phenylalanine during their lifetime suggest the need to maintain the diet. The use of DNA for diagnostic and prognostic purposes might assist in decisions about dietary quality and duration, and in anticipation of psychopathology.