We investigated the relationship between prenatal maternal urinary concentrations of phthalate metabolites and neonatal behavior in their 295 children enrolled in a multiethnic birth cohort between 1998 and 2002 at the Mount Sinai School of Medicine in New York City. Trained examiners administered the Brazelton Neonatal Behavioral Assessment Scale (BNBAS) to children within 5 days of delivery. We measured metabolites of 7 phthalate esters in maternal urine that was collected between 25 and 40 weeks' gestation. All but two phthalate metabolites were over 95% detectable. We summed metabolites on a molar basis into low and high molecular weight phthalates. We hypothesized the existence of sex-specific effects from phthalate exposure a priori given the hormonal activity of these chemicals. Overall we found few associations between individual phthalate metabolites or their molar sums and most of the BNBAS domains. However, we observed significant sex-phthalate metabolite interactions (p<0.10) for the Orientation and Motor domains and the overall Quality of Alertness score. Among girls, there was a significant linear decline in adjusted mean Orientation score with increasing urinary concentrations of high molecular weight phthalate metabolites (B=-0.37, p=0.02). Likewise, there was a strong linear decline in their adjusted mean Quality of Alertness score (B=-0.48, p<0.01). In addition, boys and girls demonstrated opposite patterns of association between low and high molecular weight phthalate metabolite concentrations and motor performance, with some indication of improved motor performance with increasing concentration of low molecular weight phthalate metabolites among boys. This is the first study to report an association between prenatal phthalate exposure and neurological effects in humans or animals, and as such requires replication.