Neurobehavioral evaluation of the high-risk neonate represents an important advance in early detection of behavioral anomalies which may give rise to later neuropsychological sequelae. In the present study neonates comprising three diagnostic categories (i.e., respiratory distressed, seizure-disordered, normals) were evaluated with the Brazelton Neonatal Behavior Assessment Scale (BNBAS) to determine the extent to which differences in neurobehavioral organization could be detected with the scale, and how they were related to diagnostic classification. Average conceptional age at testing for the three groups was within the range usually considered full term: e.g., 38.81 weeks (respiratory distressed), 40.18 weeks (normal healthy) and 42.54 weeks (seizure disorder). Infants who had been diagnosed with neonatal seizures exhibited consistently less optimal behavior than did either of the other two groups. Infants with respiratory distress and normal controls did not differ significantly on most summary measures of neurobehavioral organization scored with the BNBAS. The study offers support for the discriminative validity of the BNBAS and its potential usefulness in the assessment of clinically ill newborns.