OBJECTIVE: To assess the inter-tester and test-retest reliability and validity after developing of Korean version of National Institutes of Health Stroke Scale (NIHSS). METHOD: This research was conducted on 27 patents with stroke with less than 12 months since the onset of disease. Five physiatrists translated NIHSS into Korean. Video taping were used for objective scorings. Four physiatrists conducted scorings in order to seek for inter-tester reliability and one conducted scorings three weeks interval for test-retest reliability. Six physiatrists conducted scorings in order to seek for concurrent validity with the original NIHSS and four conducted scorings for validity with other impairment scale. Each score was analyzed based on Spear-man correlation coefficient. RESULTS: According to inter-tester reliability for Korean version of NIHSS, rho value reached over 0.70, with over 0.72 concerning test-retest reliability. The test on concurrent validity with the original NIHSS reached over 0.70 at rho value, with over 0.653 for MMSE, Motricity index, Brunnstrom stage. CONCLUSION: Newly developed Korean version of NIHSS showed high inter-tester and test-retest reliabilities, together with high concurrent validity with the original and other impairment scales, to be regarded to be used as primary impairment scale for patients with stroke.