The Lewy body composite risk score (LBCRS) is a useful clinical screening tool to help determine whether the dementia is related to Lewy body pathology. The purpose of this study is to verify reliability, validity, and diagnostic usefulness of Korean version of LBCRS (K-LBCRS). CDR-sum of boxes, Mini-Mental State Examination, and standardized scales related to cognition, mood, behavior, and motor function were administered to a total of 107 subjects, including 30 dementia with Lewy bodies (DLB), 54 Alzheimer's disease (AD), and 23 cognitively normal elderly people and their collateral informants. Internal consistency of the K-LBCRS was good with Cronbach's alpha of 0.85, and concurrent validity was also satisfactory, with K-LBCRS correlating highly with CDR-SB and other scales. The test-retest reliability was very high with a Pearson correlation coefficient of 0.97. The mean scores of K-LBCRS were significantly different among three groups, with DLB (6.2±2.4), AD (1.4±1.3), and controls (0.3±0.6). We identified a cut-off score of 3 as best to differentiate between DLB and AD, having AUC of 0.97 (95% CI 0.94-1.00), sensitivity 97%, specificity 83%, positive predictive value 76%, negative predictive value 98%, which is the same score suggested in the original study. This study shows K-LBCRS as a new useful screening tool for Korean DLB patients in clinical settings.