We prospectively examined the effect of arundic acid (AA; ONO-2506) on S-100beta, an astrocyte-derived protein, in a phase I acute stroke study. Subjects with acute ischemic stroke were randomized to daily infusion of AA or placebo for 7 days. Serum S-100beta levels were assayed pre-infusion on Days 1-7 and post-infusion on Days 1, 3, and 7, and correlated with National Institutes of Health Stroke Scale (NIHSS). Samples were obtained from 86 subjects (46 AA, 40 placebo). Increase in S-100beta protein level from baseline was less in the AA cohort than in the placebo cohort at 7 (p=0.0471; t-test) and 12 (p=0.0095)-hours post-infusion on Day 3. Baseline NIHSS correlated with maximal S-100beta levels between Days 1 and 3 in the AA (r=0.51; p=0.0003) and placebo (r=0.41; p=0.0084) cohorts and in the pooled aggregate (n=86; r=0.46; p<0.0001). The same correlations were observed between Day 10 NIHSS and Day 1-3 maximum serum S-100beta levels. Treatment with AA was associated with lower serum levels of S-100beta after acute ischemic stroke. Our results showing correlation between S-100beta and NIHSS indicate that this protein is a clinically relevant marker of neurological deficit in acute stroke.