Latino children experience high rates of delayed autism spectrum disorder (ASD) diagnosis and treatment. As primary care providers are in frequent contact with families during early childhood, delays in ASD care for Latinos may relate to providers’ practice patterns and views about autism identification. For instance, despite recommendations that pediatricians screen all children for ASD, many pediatricians do not offer ASD screening in Spanish, and most pediatricians view ASD as harder to identify in Latino children. To address these barriers, this project will: 1) Modify an existing ASD education and quality improvement program for use in pediatric practices serving low-income, bilingual Latino communities. 2) Test the program’s effectiveness in improving early autism screening and Early Intervention (autism and developmental therapy) referral rates, by comparing ASD screening and referral rates in practices that receive the program compared to those providing routine care. This study takes advantage of an existing primary care practice intervention designed to boost ASD screening and referral rates in Oregon. This award will allow program modifications, adding cultural competency elements and bilingual patient materials to make it useful for practices treating underserved Latino patients, and will also allow standardization of the program for research measurement. The improved program will be called “REAL-START” (Risk Evaluation of Autism in Latinos– Screening Tools and Referral Training). The investigators will test the effectiveness of REAL-START by measuring these primary outcome questions: 1) Whether practices receiving the program have higher rates of recommended ASD screening and less Latino/white differences in screening rates, compared to similar practices that did not receive the program. 2) Whether practices receiving the program have higher rates of referral to Early Intervention and less Latino/white differences in Early Intervention referral rates, compared to similar practices that did not receive the program. The program will be disseminated throughout Oregon and used as a model for other statewide ASD collaboratives. All project materials will be made publicly available so that the program can be replicated in other states. The study’s findings will be used as the basis for a larger federally-funded study comparing effectiveness of different types of early ASD identification.