Data from a population-based, multisite surveillance network were used to determine the prevalence of autism spectrum disorders (ASDs) among children aged 8 years in 14 areas of the United States and to describe the characteristics of these children. 2002. Children aged 8 years were identified as having an ASD through screening and abstraction of evaluation records at health facilities for all 14 sites and through information from psychoeducational evaluations for special education services for 10 of the 14 sites. Case status was determined through clinician review of data abstracted from the records. Children whose parent(s) or legal guardian(s) resided in the respective areas in 2002 and whose records documented behaviors consistent with the Diagnostic and Statistical Manual, Fourth Edition, Text Revision (DSM-IV-TR) criteria for autistic disorder; pervasive developmental disorder, not otherwise specified; or Asperger disorder were classified as having ASDs. For 2002, of 407,578 children aged 8 years in the 14 surveillance areas, 2,685 (0.66%) were identified as having an ASD. ASD prevalence per 1,000 children aged 8 years ranged from 3.3 (Alabama) to 10.6 (New Jersey), with the majority of sites ranging from 5.2 to 7.6 (overall mean: 6.6 [i.e., one of every 152 children across all sites). ASD prevalence was significantly lower than all other sites in Alabama (p<0.001) and higher in New Jersey (p<0.0001). ASD prevalence varied by identification source, with higher average prevalence for ASDs in sites with access to health and education records (mean: 7.2) compared with sites with health records only (mean: 5.1). Five sites identified a higher prevalence of ASDs for non-Hispanic white children than for non-Hispanic black children. The ratio of males to females ranged from 3.4:1.0 in Maryland, South Carolina, and Wisconsin to 6.5:1.0 in Utah. The majority of children were receiving special education services at age 8 years and had a documented history of concerns regarding their development before age 3 years. However, the median age of earliest documented ASD diagnosis was much later (range: 49 months [Utah]--66 months [Alabama]). The proportion of children with characteristics consistent with the criteria for an ASD classification who had a previously documented ASD classification varied across sites. In the majority of sites, females with an ASD were more likely than males to have cognitive impairment. For the six sites for which prevalence data were available from both 2000 and 2002, ASD prevalence was stable in four sites and increased in two sites (17% in Georgia and 39% in West Virginia). Results from the second report of a U.S. multisite collaboration to monitor ASD prevalence demonstrated consistency of prevalence in the majority of sites, with variation in two sites. Prevalence was stable in the majority of sites for which 2 years of data were available, but an increase in West Virginia and a trend toward an increase in Georgia indicate the need for ongoing monitoring of ASD prevalence. These ASD prevalence data provide the most complete information on the prevalence of the ASDs in the United States to date. The data confirm that ASD prevalence is a continuing urgent public health concern affecting an approximate average of one child in every 150 and that efforts are needed to improve early identification of ASDs.