Elopement is a common and dangerous behavior among autistic children. Behavioral treatments can reduce elopement, but most evidence comes from small-N evaluations in specialized settings with strategies varying across studies. The current study compared the efficacy of the caregiver-mediated function-based elopement treatment to parent education program (PEP) in a 16-week randomized clinical trial of 76 autistic children (age = 4-12 years). Function-based elopement treatment involves 12 weekly appointments aimed at improving safety, identifying the function of elopement, and implementing subsequent function-based treatment strategies. No group differences were observed on the Aberrant Behavior Checklist-Hyperactivity (primary outcome). Significant improvement from baseline to endpoint in function-based elopement treatment compared to parent education program participants was observed for secondary outcomes, including caregiver ratings of safety measures (p < 0.01), severity of elopement based on the Elopement Questionnaire (p < 0.01), and caregiver-collected data on elopement (p < 0.01). The Clinical Global Impression--Improvement Scale (CGI-I) rated by a treatment-blind evaluator found 31.6% of function-based elopement treatment participants improved compared to 2.6% in parent education program (p = 0.001). Improvements were maintained at a 28-week follow-up. Attrition was 5.26%, and no significant adverse events were deemed related to treatment. Function-based elopement treatment was superior to parent education program on elopement-specific outcomes and appears safe and acceptable.