Objective: Most tests of cognitive behavioral therapy (CBT) for youth anxiety disorders have shown beneficial effects, but these have been efficacy trials with recruited youths treated by researcher-employed therapists. One previous (nonrandomized) trial in community clinics found that CBT did not outperform usual care (UC). The present study used a more stringent effectiveness design to test CBT versus UC in youths referred to community clinics, with all treatment provided by therapists employed in the clinics. Method: A randomized controlled trial methodology was used. Therapists were randomized to training and supervision in the Coping Cat CBT program or UC. Forty-eight youths (56% girls, 8 to 15 years of age, 38% Caucasian, 33% Latino, 15% African-American) diagnosed with "DSM-IV" anxiety disorders were randomized to CBT or UC. Results: At the end of treatment more than half the youths no longer met criteria for their primary anxiety disorder, but the groups did not differ significantly on symptom (e.g., parent report, eta-square = 0.0001; child report, eta-square = 0.09; both differences favoring UC) or diagnostic (CBT, 66.7% without primary diagnosis; UC, 73.7%; odds ratio 0.71) outcomes. No differences were found with regard to outcomes of comorbid conditions, treatment duration, or costs. However, youths receiving CBT used fewer additional services than UC youths (chi[superscript 2[subscript 1]] = 8.82, p = 0.006). Conclusions: CBT did not produce better clinical outcomes than usual community clinic care. This initial test involved a relatively modest sample size; more research is needed to clarify whether there are conditions under which CBT can produce better clinical outcomes than usual clinical care. (Contains 1 figure and 1 table.)