Objective This study aims to examine the real‐world effectiveness of a computer‐assisted cognitive behavioral therapy (CBT) protocol relative to treatment as usual (TAU) among anxious children presenting at community mental health centers. Methods One hundred children (7–13 years) with clinically significant anxiety were randomized to receive either 12 weekly computer‐assisted CBT sessions or TAU for an equivalent duration. Assessments were conducted by independent evaluators at screening/baseline, midtreatment, posttreatment, and 1‐month followup (for computer‐assisted CBT treatment responders). Results There were significant between‐group effects favoring the computer‐assisted CBT condition on primary anxiety outcomes. Thirty of 49 (61.2%) children randomized to computer‐assisted CBT responded to treatment, which was superior to TAU (6/51, 11.8%). Relative to TAU, computer‐assisted CBT was associated with greater reductions in parent‐rated child impairment and internalizing symptoms, but not child‐rated impairment and anxiety and depressive symptoms. Treatment satisfaction and therapeutic alliance in those receiving computer‐assisted CBT was high. Treatment gains in computer‐assisted CBT responders were maintained at 1‐month followup. Conclusions Within the limitations of this study, computer‐assisted CBT is an effective and feasible treatment for anxious children when used in community mental health centers by CBT‐naïve clinicians.