Online acceptance and commitment therapy (ACT) is promising for treating a range of psychological problems. Component research can further clarify which components are needed for optimal outcomes in what contexts. Online platforms provide a highly controlled format for such research. In this pilot trial, 55 adults were randomized to: ACT-Open (i.e., acceptance, defusion components), ACT-Engaged (i.e., values, committed action), or ACT-Combined (i.e., acceptance, defusion, values, committed action). Each condition was 12 sessions over six weeks, with assessments at baseline, posttreatment, and four-week follow-up. ACT-Open, ACT-Engaged, and ACT-Combined all significantly improved from pre- to post-treatment on mental health, psychosocial functioning, and components of psychological flexibility. Compared to ACT-Combined, ACT-Open improved less on psychosocial functioning at posttreatment, and ACT-Engaged worsened on functioning at follow-up. The platform was acceptable with high satisfaction ratings. Results support the feasibility of conducting online ACT component research, which will be tested in a fully powered non-inferiority trial.