Clinical observations over a 36-month period on a group of patients with phenylketonuria are summarized. Preliminary comparison of 9 phenylketonuric patients started on dietary treatment at a mean age of 3.6 weeks and treated for a mean duration of 13.5 months revealed significantly lower developmental achievement than in their 34 unaffected siblings. Attempts should be made to shorten the period between birth and exact diagnosis as much as possible in order to determine whether that will further increase the benefits of dietary therapy. Maternal obstetrical histories were found to be significantly abnormal in approximately 46% of net pregnancies in heterozygotes. Consequently, careful evaluation of the role of perinatal factors is advised in the assessment of the results of dietary therapy in each patient. Fasting serum phenylalanine levels of 16 untreated infants under a month of age were not significantly different from those in 20 older patients with PKU. Individual variations in the vulnerability of the nervous system to elevation of the blood phenylalanine level may be another important factor in the variability in intellectual defect. One atypical patient is discussed. Measurements of phenylalanine levels in blood and close observation of the clinical course of the patient, particularly his weight gain, rather than the presence or absence of abnormal metabolites of phenylalanine in the urine, should be used as guides to early diagnosis and to the dietary management of patients with phenylketonuria.