Objectives : The purpose of this study was to investigate the possibility of Bi-Su as a pattern identification (PI) index in stroke patients. Methods : The subjects were 424 hospitalized stroke patients within 1 month from onset and diagnosed with the same PI subtypes (dampness & phlegm, qi deficiency, fire & heat, eum deficiency, and blood stasis) by agreement of two clinical experts. Bi-Su type is a kind of body shape (Bi : fat, Su : lean). Bi-Su type and degree (Bi-Su score) were decided by clinical expert. Body mass index (BMI) and waist-hip ratio (WHR) were used as an obesity index. Correlation analysis between Bi-Su score and obesity index (Spearman) and variance analysis for Bi-Su score, BMI, and WHR among PI subtypes (ANOVA) and sex were carried out. Results : While there was partial correlation between Bi-Su type and BMI($r^2$=0.634, p<0.001), the distribution of the BMI group based on the Bi-Su group showed the broadest range. The Bi-Su score in the dampness & phlegm group was higher than in the other groups (p<0.001). BMI in the dampness & phlegm groups was also higher but the BMI differences among PI subtypes was low (p=0.002). The Bi-Su score in the dampness & phlegm group was similar in both sexes, although the hand score in the eum deficiency group was the lowest, especially in males. Conclusions : Although BMI is not an objective enough tool for evaluating Bi-Su type, Bi-Su type is more appropriate than BMI as PI index. Therefore Bi-Su type could be used as one of the PI indices for dampness & phlegm or eum deficiency group in stroke patients.