This research aimed to establish whether 12 weeks of bicycle-riding skills exercises with massed and distributed practice frequency at low and high intensity affect communication performance and stereotyped behavior among adolescent boys with autism. Fifty autistic boys aged 13.3 ± 1.32 years participated in the study. The participants were divided into homogeneous experimental groups (N = 10) with dissimilar training frequencies and intensities, along with a control group. Pre-tests using the GARS-2 test were administered to assess stereotyped behavior and communication skills, followed by post-tests and a one-month follow-up. Non-parametric Kruskal Wallis and ANCOVA results at a significance level of 0.05 showed that there was a significant difference in the post-test of stereotyped behavior and communication skills (p = 0.001 and p = 0.002, respectively) and follow-up test one month later (p = 0.003, p = 0.048, respectively) between the intervention and control groups after performing bike riding skills exercises with low and high intensities and frequencies (one and three sessions per week). Regardless of the intensity and frequency, bike riding skills training during the critical period of adolescence can significantly reduce stereotyped behaviors and enhance communication skills, which can also support positive development in other domains for individuals with autism.