Engaging in spontaneous social behaviors towards others is critical to initiating and maintaining reciprocal social interactions (e.g., Garner & Estep, 2001; Oke & Schreibman, 1990). The literature has identified marked deficits in child-initiated social behaviors in young children with autism (e.g., Koegel, Koegel, Frea, & Fredeen, 2001; Mundy & Burnette, 2005). Researchers believe that interventions effectively targeting these early social communication behaviors in this population may minimize obstacles to subsequent language learning and social interaction skills (e.g., Rogers & Dawson, 2010; Yoder, Warren, & Hull, 1995). Thus far, there is very little research on specific behavioral training of social initiations to young, preverbal children with autism or on the effect this early training has on response to treatment. The aim of this research project was to systematically evaluate a social initiation training (SIT) program for children with autism and examine the effect of SIT on the efficacy of another behavioral treatment, PRT. This study targeted children who, based on an assessment of specific behavior characteristics, were not expected to increase social initiation behaviors during PRT (Sherer & Schreibman, 2005). A single-subject multiple baseline design across subjects was used to examine treatment efficacy and to evaluate individual differences in treatment response. During baseline sessions children received PRT only. During the treatment component of the study, children's sessions included both SIT and PRT. This study replicated earlier findings that children who do not exhibit the PRT "responder" profile do not increase social initiations in response to PRT alone. As hypothesized, with the addition of an SIT program, all four participants evidenced gains in social initiations. All four children displayed gains in vocal communication during treatment (PRT + SIT) that were greater than gains in baseline (PRT). The three participants who did not display approach behaviors consistent with PRT "responders" profile prior to treatment, met the criteria following treatment. These results support the notion that at least one of the behaviors in the PRT "responder" profile, approach, can change in response to therapeutic intervention. Implications for the importance of social initiation training with young preverbal children are discussed