More than ten years of research has been conducted on the use of social stories to change the behavior of children with autism, yet the results of these studies continue to be inconclusive. While many articles indicate that social stories do result in positive behavior change, these articles are fraught with errors or threats to internal validity. The most promising results to date indicate that social stories are most effective when combined with prompts and/or reinforcement. This study sought to examine the impact of social stories on independent and prompted peer-directed social interactions while measuring frequencies of prompts and reinforcers available in the natural environment. These data were then compared to results obtained when prompts were combined with social stories or prompts were delivered without social stories, continuing to collect data on incidence of positive consequences. These data were evaluated using modified reversal designs with four participants who had all been diagnosed with autism. Three males and one female, with ages ranging from 10 to 21 years old, served as participants. Results indicate that increases over baseline in independent appropriate interactions were seen for all four participants, especially during the social story plus prompts condition and prompts only conditions. With only one exception, the social story alone (without prompts) did not result in increased social interaction. The greatest increases occurred when natural positive consequences were available at high rates. These data provide direction for future research on the use of social stories including incorporating prompts into the social story protocol and scheduling reinforcement for behaviors identified in the social story.