Objective: Explore the associations between the tongue appearances in Traditional Chinese Medicine (TCM) and effective response (ACR20 response based on American College of Rheumatology) in rheumatoid arthritis (RA) patients treated with Chinese medicine (CM) and western biomedical combination therapy (WM). Methods: This study used the data from a previous multi-center randomized-controlled clinical trial. Data pertaining to tongue coating and tongue body color were collected. In order to simplify the tongue diagnosis for easily understood by biomedical professionals, only two typical tongue coating (white and yellow) and four typical tongue body colors (purple, pink, pale and red) were identified for this analysis. 170 cases with clear tongue coating and 198 cases with clear tongue body color in TCM treatment (Glucosidorum Tripterygll Totorum tablets and Yishen Juanbi tablets) group, 181 cases with identified tongue coating and 189 cases with identified tongue body color in WM treatment (diclofenec, methotrexate and sulfasalazine) group were included for the analysis. The ACR20 response at 12 weeks and 24 weeks were used as an outcome measure of efficacy. The effective rates in patients with different tongue appearances were analyzed with Chi-square method and the association between the changes of tongue coating/body color and the ACR20 response was analyzed with a repeated measures logistic regression analysis. Results: At 12 weeks, the ACR20 responses in the patients treated with CM and WM therapy were 33.6% and 53.0%, respectively, and at 24 weeks, they were 57.9% and 84.3%, respectively. RA patients with white tongue coating showed higher effective rate than those patients with yellow tongue coating in the treatment with WM intervention (p < 0.05), and there was no difference in the patients with CM intervention. Further association analysis showed that TCM would be less effective for the patients with pale tongue body (p = 0.0323), and WM would be less effective for the patients with purple or red tongue body (p = 0.0291 and 0.0027, respectively). Conclusion: TCM was less effective for the patients with pale tongue body, and WM was be less effective for the patients with purple or red tongue body, or white tongue coating. The results suggest that tongue coating and body color might be used to help identify a subset of RA patients both for CM and WM interventions. (C) 2011 Elsevier Ltd. All rights reserved.