It was suggested that a ''fetal cocaine syndrome'' exists. The objective of this study was to systematically investigate whether or not a ''cocaine syndrome'' exists. The setting was Parkland Memorial Hospital, a large urban public hospital in Dallas, TX, where approximately 15,000 infants are delivered annually. Infants who tested positive by urinalysis for cocaine (n = 25) were included in this study. Controls negative for cocaine (n = 25) were matched to cocaine-exposed infants for estimated gestational age, sex, and race. A standardized dysmorphology examination (135 features) and a series of anthropometric measures (n = 22) were done for each cocaine-exposed and control infant by an observer blinded to drug-exposure status. Fetal growth retardation characterized cocaine-exposed infants. No characteristic pattern of minor dysmorphic or anthropometric features of the face, limbs, or torso was observed among cocaine-exposed infants. Cocaine-exposed infants lack a facial gestalt or torso/limb features that would characterize a syndrome. If a ''cocaine syndrome'' that can be characterized dysmorphologically and/or anthropometrically exists, its occurrence seems infrequent. (C) 1996 Wiley-Liss, Inc.