Objective To determine whether application of criteria for remission in rheumatoid arthritis (RA) may result in underestimation of foot joint involvement among patients in a clinic setting. Methods RA patients (n = 123) were assessed at baseline and 6 months after commencement of a response-driven combination disease-modifying antirheumatic drug (DMARD) protocol. Remission was assessed using disease activity measures (the 28-joint Disease Activity Score using the erythrocyte sedimentation rate [DAS28-ESR], Simplified Disease Activity Index [SDAI], and Clinical Disease Activity Index [CDAI]) as well as Boolean-based criteria for remission (the 1981 American College of Rheumatology [ACR] preliminary criteria and the 2011 ACR/European League Against Rheumatism [EULAR] provisional criteria). The prevalence of foot synovitis and the mean swollen/tender foot joint count in RA patients meeting any of these remission criteria were estimated by hurdle (mixed distribution) regression. Results In patients who received 6 months of combination DMARD treatment, application of the 1981 ACR criteria and the newly proposed 2011 ACR/EULAR criteria, each utilizing full joint counts (which includes assessment of the feet), classified the least number of patients as being in remission (8-10%), and evidence of foot synovitis was minimal among these patients. In contrast, ongoing foot synovitis was present in a substantial proportion of patients (>20%) meeting the 28-joint count criteria for remission, including the DAS28-ESR, SDAI, CDAI, and 2011 ACR/EULAR criteria (clinical practice setting or clinical trials). Furthermore, applying the 2011 ACR/EULAR composite remission criterion of a SDAI score 3.3 to define remission did not adequately capture the resolution of foot synovitis (i.e., residual foot involvement was still detected in a substantial proportion of patients classified as being in remission by this definition). Conclusion Although the DAS28-ESR, CDAI, and SDAI have been validated for assessment of remission in RA, this study shows that the performance of these 3 disease activity measures, which do not provide a direct assessment of the foot, in detecting foot synovitis is poor, in contrast to that of the 1981 ACR and 2011 ACR/EULAR remission criteria utilizing full joint counts. Thus, patients may be at risk of ongoing damage if treatment decisions are made solely on the basis of criteria that omit foot joint assessment.