OBJECTIVE To assess the prevalence of posttraumatic stress reactions (PTSR) in New York City (NYC) children following the September 11, 2001, attacks and determine the key predictors of PTSR. METHODS Cross-sectional random digit-dial survey in NYC of parents of children 4-17 years old 4 months after the attacks. PTSR in children was measured using the 20-item Posttraumatic Stress Disorder Reaction Index-Child Revision, with parents as respondents. RESULTS Overall, 18% of NYC children had "severe" or "very severe" PTSR, and 66% had "moderate" PTSR. In a multivariate model, parental posttraumatic stress disorder (PTSD; odds ratio [OR] = 4.50; P <.01), the parent crying in front of the child (OR = 3.19; P <.001), seeing 3 or more graphic images of the disaster on television (OR = 3.18; P <.01), and living in Manhattan were associated with severe or very severe PTSR in children. CONCLUSIONS A substantial proportion of NYC children had severe or very severe PTSR after September 11, and most children exhibited at least moderate PTSR. These findings suggest an enhanced role for primary care physicians, particularly pediatricians, for screening, treatment, and referral (coupled with appropriate training and reimbursement), especially in light of continued terrorist threats. These findings also have implications for advice that pediatricians can give to parents about limiting disaster-related television exposure and children's need for emotional support.