The literature and the results of our previous finding periods have supported the position that parent training is an extremely effective approach to the treatment of autism. It is fast, economical, and produces relatively broad treatment changes. During our research program our work has been concerned with identifying specific variables related to the best-practice parent training and developing an improved program designed to address these variables. Accordingly, our research has assured a progressive direction in the development of a parent training treatment delivery package that is optimal for the child and for the family. In our previous research we have typically compared one type of treatment to another type of treatment. Looking back at our efforts in this area we now feel we have a substantial corpus of data both comprehensive in scope and rich in detail; and based on our pilot studies and preliminary investigations we believe that careful analyses of these data now allow us to specify several variables important for the purpose of developing "individualized" treatments for children with autism and their parents. Our data indicate that although we have consistently found one standard type of treatment to be more beneficial overall than another standard type of treatment, each individual treatment contains aspects that are very powerful for specific purposes. In fact, our data suggest that a combination of these treatment procedures, individualized based on child, family and target behavior characteristics will be far superior than implementing one type of treatment for all children as if they were all the same. Because of the large amount of heterogeneity in child and parent characteristics seen in the area of autism it is becoming abundantly clear that individualized treatment will be greatly superior to a standard package. We hypothesize that this type of treatment will significantly improve the treatment of autism, with respect to both direct measures of child behavior and with respect to measures of overall family functioning. We thus propose to carefully compare tow treatment conditions. Our control condition will be our current best-practice parent training package (the Self-Management condition, now called the Standard Package or SP condition). Our experimental condition will be one in which treatment will be "individualized" for each family in that the specifies of the treatment plan will be dictated by the characteristics of the child and of the family (the Individualized Package or IP condition).