The prevalence of Autism Spectrum Disorder (ASD) has increased dramatically in recent years, with recent data reporting one out of every 150 children being affected (CDC MMWR, 2007). Although our understanding of the cause, course, and prognosis of ASD has developed greatly in past decades, there has been a significant lack of research focusing on sociodemographic factors that may influence rates of identification and subsequent treatment. This is of particular concern because significant social disparities in autism services have been identified (Mandell, et al., 2009). Families of racial/ethnic minority, lower levels of education, and those who live in non-metropolitan areas experience greater limitations in accessing services for ASD (Thomas, Ellis, McLaurin, Daniels, & Morrissey, 2007). The purpose of the proposed project is to expand, adapt, and evaluate a modified version of an early intervention working with caregivers from a low-income, underserved population. This study proposes a randomized group crossover experimental design with 60 children randomly assigned to receive either a caregiver-implemented intervention or community support. The study aims 1) to compare of the effectiveness of the two treatment conditions on outcome measures of social communication skills, autism symptoms, developmental levels, and adaptive behavior; 2) to identify specific child and family characteristics that predict response to intervention, as well as providing a preliminary examination of how factors at the caregiver level mediate child outcome measures; and 3) to test the overall applicability of the early intervention in the targeted population. The study will begin to investigate moderating and mediating factors of service utilization, satisfaction, and adherences.