Objective To obtain the effective characteristic information for the standardization of diagnosis of traditional Chinese medical(TCM) syndrome in rheumatoid arthritis(RA).Methods Fifty-four items of symptoms and signs in RA patients were input into the computer,and Mahalanobis distance discriminant analysis(DDA) was used for the selection of effective indexes.Results According to the weight value and the result of T test,17 items from 54 had the highest rate of discrimination accuracy,up to 96.5%.Conclusion The results of computerized pattern recognition for the syndrome diagnosis of RA are consistent with the clinical diagnosis,which indicates that computerized pattern recognition can be used for the standardization of syndrome diagnosis of RA.