Background Behavioral interventions are commonplace in the treatment of autism spectrum disorders, yet relatively little is known about how and why these interventions work. This study tests the relationship between isolated core components of a packaged social communication intervention and the primary outcome, joint engagement, to better understand how the intervention is affecting change in individuals. Methods A total of 86 toddlers and their parents were enrolled in the study and randomized to one of two treatments, the joint attention, symbolic play, engagement, and regulation (JASPER) parent-mediated intervention or a psychoeducational intervention. Measures regarding the parent's use of intervention strategies were collected before and after the 10-week intervention. Additional measures of child and parent joint engagement were also collected. Results A significant effect of treatment was found for all four of the core strategies of the intervention, favoring a larger increase in the JASPER condition. A hierarchical linear regression revealed several individual predictors of joint engagement, including parent-rated buy-in, interventionist-rated parent involvement, and parental use of strategies. To complement the hierarchical analysis, we also tested the potential mediating effect the strategies may have on the relationship between treatment and joint engagement. Results showed that the strategy of mirrored pacing mediated the relationship between treatment and joint engagement in the positive direction. Conclusions These results strongly suggest that the mirrored pacing strategy is an active ingredient of the JASPER treatment.