Purpose Fear of cancer recurrence (FCR) is prevalent, distressing, and long lasting. This study evaluated the impact of a theoretically/empirically based intervention (ConquerFear) on FCR. Methods Eligible survivors had curable breast or colorectal cancer or melanoma, had completed treatment (not including endocrine therapy) 2 months to 5 years previously, were age > 18 years, and had scores above the clinical cutoff on the FCR Inventory (FCRI) severity subscale at screening. Participants were randomly assigned at a one-to-one ratio to either five face-to-face sessions of ConquerFear (attention training, metacognitions, acceptance/mindfulness, screening behavior, and values-based goal setting) or an attention control (Taking-it-Easy relaxation therapy). Participants completed questionnaires at baseline (T0), immediately post-therapy (T1), and 3 (T2) and 6 months (T3) later. The primary outcome was FCRI total score. Results Of 704 potentially eligible survivors from 17 sites and two online databases, 533 were contactable, of whom 222 (42%) consented; 121 were randomly assigned to intervention and 101 to control. Study arms were equivalent at baseline on all measured characteristics. ConquerFear participants had clinically and statistically greater improvements than control participants from T0 to T1 on FCRI total ( P < .001) and severity subscale scores ( P = .001), which were maintained at T2 ( P = .017 and P = .023, respectively) and, for FCRI total only, at T3 ( P = .018), and from T0 to T1 on three FCRI subscales (coping, psychological distress, and triggers) as well as in general anxiety, cancer-specific distress (total), and mental quality of life and metacognitions (total). Differences in FCRI psychological distress and cancer-specific distress (total) remained significantly different at T3. Conclusion This randomized trial demonstrated efficacy of ConquerFear compared with attention control (Taking-it-Easy) in reduction of FCRI total scores immediately post-therapy and 3 and 6 months later and in many secondary outcomes immediately post-therapy. Cancer-specific distress (total) remained more improved at 3- and 6-month follow-up.