Background Mental illness stigma has been a global concern, owing to its adverse effects on the recovery of people with mental illness, and may delay help-seeking for mental health because of the concern of being stigmatized. With technological advancement, internet-based interventions for the reduction of mental illness stigma have been developed, and these effects have been promising. Objective This study aimed to examine the differential effects of internet-based storytelling programs, which varied in the levels of interactivity and stigma content, in reducing mental illness stigma. Methods Using an experimental design, this study compared the effects of 4 storytelling websites that varied in the levels of interactivity and stigma content. Specifically, the conditions included an interactive website with stigma-related content (combo condition), a noninteractive website with stigma-related content (stigma condition), an interactive website without stigma-related content (interact condition), and a noninteractive website without stigma-related content (control condition). Participants were recruited via mass emails to all students and staff of a public university and via social networking sites. Eligible participants were randomized into the following four conditions: combo (n=67), stigma (n=65), interact (n=64), or control (n=67). The participants of each group viewed the respective web pages at their own pace. Public stigma, microaggression, and social distance were measured on the web before the experiment, after the experiment, and at the 1-week follow-up. Perceived autonomy and immersiveness, as mediators, were assessed after the experiment. Results Both the combo (n=66) and stigma (n=65) conditions were effective in reducing public stigma and microaggression toward people with mental illness after the experiment and at the 1-week follow-up. However, none of the conditions had significant time×condition effects in reducing the social distance from people with mental illness. The interact condition (n=64) significantly reduced public stigma after the experiment (P=.02) but not at the 1-week follow-up (P=.22). The control condition (n=67) did not significantly reduce all outcomes associated with mental illness stigma. Perceived autonomy was found to mediate the effect of public stigma (P=.56), and immersiveness mediated the effect of microaggression (P=.99). Conclusions Internet-based storytelling programs with stigma-related content and interactivity elicited the largest effects in stigma reduction, including reductions in public stigma and microaggression, although only its difference with internet-based storytelling programs with stigma-related content was not statistically significant. In other words, although interactivity could strengthen the stigma reduction effect, stigma-related content was more critical than interactivity in reducing stigma. Future stigma reduction efforts should prioritize the production of effective stigma content on their web pages, followed by considering the value of incorporating interactivity in future internet-based storytelling programs. Trial Registration ClinicalTrials.gov NCT05333848; https://clinicaltrials.gov/ct2/show/NCT05333848