Leptin has a close correlation with obesity, which is known to be a major factor for stroke. This study was performed to determine whether serum leptin level would be an independent risk factor for stroke and whether it would change significantly early after stroke. Subjects were selected from those within I month after onset and non-stroke referents at Kyung Hee Oriental Medical Center in Seoul, Korea. We compared leptin and the other characteristics between stroke subjects and referents. Body mass index, hypertension history, presence of drinking and smoking, waist/hip ratio, total cholesterol and triglyceride were recorded. To assess odds ratio of leptin for stroke, we used logistic regression analysis. Leptin was rechecked 2 weeks later and compared with the former value in acute stroke subjects. In this study, serum leptin did not differ significantly between stroke subjects and referents, and its odds ratio was not significant in male (OR=0.52, 95％ Cl; 0.13-2.08) and female (OR=1.57, 95％ Cl; 0.53-4.67). In acute stroke subjects, leptin did not change significantly 2 weeks later. Hypertension had a significant odds ratio in male (OR=3.39, 95％ Cl; 1.02-11.24) and female (OR=12.37, 95％ Cl; 3.67-41.65). High waist/hip ratio was only in female (OR=6.70, 95％ Cl; 1.73-26.02). In conclusion, leptin was not an independent risk factor for stroke and its serum level did not change significantly early after stroke. Hypertension and waist/hip ratio had significant relative risks.