To improve the overall quality and effectiveness of mental health intervention, it is essential to identify empirically supported behavioral treatments(ESBTs) and assure the accessibility and usability of these treatments for providers in community settings. Fidelity of implementation (FI), or the degree to which an intervention is implemented as was intended, provides necessary information for training and maintaining providers' effectiveness. However, the tools and procedures used to measure FI and provide feedback to providers must be effective and efficient. Through archival data analysis and prospective assessment of individual treatment components, we aim to develop and assess an FI instrument and assessment procedures for an ESBT for children with autism. Autism is a key area of concern in the field of child mental health; Pivotal Response Training (PRT) is a naturalistic ESBT for this population (Humphries, 2003; National Standards Project, 2009). PRT is routinely used in community settings with limited data regarding treatment integrity or effectiveness (Stahmer, Collings & Palinkas, 2005). PRT fidelity of implementation (FI) measures were developed for use in research settings and have received minimal validation and adaptation for use by community therapists. The specific components of PRT have not been independently evaluated to determine how FI relates to child outcome, and the effect of ancillary factors such as therapist discipline, setting, education, affect, and the treatment environment on treatment integrity and child outcome have not been examined. The overall goal of this project is to inform strategies for enhancing fidelity of PRT in community service settings Toward this goal, there will be six specific aims: 1) Evaluate relationships between FI of PRT and within-session child skills and behavior via archival data analysis. 2) Evaluate relationships between provider characteristics and FI of PRT components via archival data analysis. 3) Individually test possible modifications to components of PRT using deconstructive evaluation. 4) Develop and evaluate a PRT FI instrument and assessment procedures in partnership with community providers. 5) Evaluate effectiveness of a PRT FI instrument and assessment procedures including (a) examining improvements to training and supervision methods using the instrument and (b) analyzing possible moderating variables affecting FI of PRT and child outcome in a prospective study. 6) Prospectively assess the relationship between provider implementation of PRT and child skills and behavior. The proposed scientific model will generate data and important products supporting ESBT use in the autism services community. This research will also serve as a general model for evaluation and improvement of FI knowledge and practice for the mental health services community and implementation science.