Birth weight has long been a focus of study by epidemiologists and human biologists, because it reflects the quality of the intrauterine environment and may be used as a predictor of future growth and development. Comparisons of Black and White neonates in the USA have consistently shown differences in birth weight. Confounding variables are a major problem in any such investigation, especially socio-economic status which is highly correlated with race in the USA. This study was distinctive in the sampling of one socio-economic stratum (low income), and the use of five anthropometric measures in addition to birth weight. The goals of this study were as follows: to determine if there were differences in body size and body composition at birth in Black and White neonates of low socio-economic status (SES), and to investigate what variables might account for any observed variability. The sample consisted of full term Black and White neonates of low SES (n = 323) born in Albany, NY (1986-1997). Birth weight, length, head and arm circumference, and subscapular and triceps skinfolds were compared. Race was determined through maternal self-identification. White neonates were significantly larger than Black neonates in birth weight, length and head circumference. Among female neonates none of the anthropometric dimensions differed between Blacks and Whites. Among male neonates, Whites were significantly larger than Blacks in birth weight, length, head and arm circumferences. Principal components analysis reduced the six anthropometric dimensions to two summary measures: body size and composition. When controlling for social and biological variables, race and sex were significant predictors of body composition, but not body size. Interpretation of results and possible causal relationships are discussed.