Objective. Ethical constraints on the conduct of placebo-controlled trials evaluating new therapies for serious chronic diseases, such as rheumatoid arthritis (RA), indicate the need for discerning methods to assess treatment effect in active-controlled clinical trials. Dynamic gadolinium-enhanced magnetic resonance imaging (DEMRI) is a sensitive technique for the detection of synovial inflammation in RA. Therefore, this investigation was undertaken to evaluate DEMRI as an efficacy assessment tool for differentiating treatment effect in a randomized, active-controlled trial comparing leflunomide and methotrexate.Methods. Patients with active RA (n ؍ 39) were randomized in a 2-center, prospective, double-blind clinical trial to receive either leflunomide (n ؍ 18) or methotrexate (n ؍ 21) therapy for 4 months. DEMRI scans were obtained at baseline and at 4 months, and the initial rate of enhancement (IRE) and the maximal signal intensity (SI) enhancement (ME) were calculated from the SI curves. Clinical improvement was assessed by conventional outcome measures.Results. Thirty-four patients (17 treated with leflunomide and 17 with methotrexate) had usable baseline and end point DEMRI scans. Leflunomide treatment was associated with a significantly greater improvement in IRE compared with methotrexate treatment (P < 0.05). Average values of ME indicated reduction of inflammation with both leflunomide and methotrexate. The improvement in clinical signs and symptoms, as measured by traditional assessments, was comparable for both active treatments.Conclusion. Results of this study validate the sensitivity of DEMRI in detecting inflammatory changes in active RA in response to treatment. Improvement in synovial inflammation as measured by IRE was significantly better with leflunomide than with methotrexate over 4 months of therapy.