Objective: We sought to investigate whether serum cystatin C levels are correlated with either stroke severity or with potential risk factors of acute ischemic stroke. Methods: 171 patients with acute ischemic stroke and 99 control subjects with minor, unrelated diseases with stroke were included in this retrospective study. Serum cystatin C levels were determined in all subjects. Serum concentrations of several vascular risk factors in stoke patients were determined by biochemical assays. The severity of strokes was scored via the National Institutes of Health Stroke Scale (NIHSS). Results: Serum cystatin C levels were significantly increased in patients with acute ischemic stroke compared with control subjects (1.26 +/- 0.34 mg/L vs. 0.78 +/- 0.24 mg/L, p < 0.001). When analyzed in quartiles of serum cystatin C levels, concentrations were low (< 0.75 mM) for 5 stroke patients (2.92%), intermediate (0.75-1 mM) for 42 patients (24.56%), high (1-1.25 mM) for 45 patients (26.32%), and very high (> 1.25 mM) for 79 patients (46.20%). However, serum cystatin C levels were not correlated with NIHSS scores, serum total cholesterol, high-density lipoprotein, low-density lipoprotein, apolipoprotein a, or apolipoprotein b levels. Further, serum cystatin C concentrations in stroke patients were not correlated with the presence of intracranial arterial stenosis, hypertension, or diabetes. Conclusion: Our study suggests that there is a close relationship between cystatin C and acute ischemic stroke, independently of conventional risk factors. But the levels of cystatin C are not correlated with the stroke severity.