To compare magnetic resonance imaging (MRI) and ultrasonography (US) in the detection of joint inflammation of rheumatoid arthritis (RA), 6 patients with RA were examined by US and low-field 0.3-T nonenhanced dedicated extremity MRI (compacTscan). All patients were females, with mean age of 50.2 years, mean disease duration of 13.5 years, and mean disease activity score (DAS)28-CRP of 1.78. Each patient was treated with either infliximab, etanercept, adalimumab, or tocilizumab. Intercarpal joints, radioulnar joints, second through fifth proximal interphalangeal (PIP) joints, and first through fifth metacarpophalangeal (MCP) joints (a total of 132 joints, 22 joints in each patient) were assessed by MRI for presence of joint inflammation. A total of 156 joints (24 first interphalangeal and radiocarpal joints plus the above 132 joints), were assessed by grayscale US (GS-US) and power Doppler US (PD-US) for presence of joint inflammation by two trained ultrasonographers. We assessed correlations between joint inflammations on MRI and GS-US/PD-US, and also interobserver correlation between the two ultrasonographers by calculating intraclass correlation coefficients (ICC). Synovial hypertrophy and/or synovial fluid was detected in 74/156 joints on GS-US, and synovitis was detected in 10/156 joints on PD-US and in 38/132 joints on MRI. Using PD-US as a reference, sensitivity of MRI in detection of synovitis was 80%. Using MRI as a reference, sensitivity of PD-US was 21%. Specificity of PD-US was higher than that of MRI. Overall agreement between GS-US and MRI and between PD-US and MRI was 0.56 and 0.76, respectively, suggesting that results of PD-US are close to those of MRI. ICC was 0.545 for GS-US and 0.807 for PD-US, suggesting specificity of PD-US in detecting joint inflammation. Our results show that findings of PD-US correlated with those of MRI. Low-field MRI and PD-US are useful tools for assessment of patients with RA.