Background Participation in electronic media use among 2-3 year olds is high and associated with adverse health and developmental outcomes. This study sought to test the feasibility and potential efficacy of a family-based program to decrease electronic media (EM) use in 2-3-year-old children. Methods Family@play was a six-session pilot randomised controlled trial delivered to parents of 2-3 year-old children from August to September 2012 in a community environment in the Illawarra region of New South Wales, Australia. Development of program content was guided by Social Cognitive and Family Systems Theories. The primary outcome was children's electronic media use. Secondary outcomes included children's time in sitting, standing and stepping. Data collectors were blinded to group allocation. Parents completed comprehensive process evaluation measures and participated in focus group discussions following completion of the program. Regression analyses were undertaken and effect sizes calculated using principles of intention to treat. Results Twenty-two participants (n = 12 intervention; n = 10 control) provided complete baseline data; complete data from 16 participants (n = 6 intervention; n = 10 control) were available post-intervention. Process evaluation results were high, showing the acceptability of the program. Compared with children in the control group, there were greater decreases in total EM use among children in the intervention group (adjusted difference [95 % CI] = −31.2 mins/day [−71.0-8.6] Cohen's d = 0.70). Differences for other outcomes were in the hypothesised direction and ranged from small for postural (sitting, standing, stepping) outcomes to moderate to large for individual electronic media (e.g. TV viewing, DVD/video viewing). Conclusions This is the first family-based study to engage families of 2-3 year old children outside the United States and target multiple EM behaviours. Family@play was shown to be a feasible and acceptable intervention to deliver to families of 2-3 year old children. Potential efficacy is evident from moderate to large effect sizes. A larger trial is warranted to test the efficacy of the program.