Background: Cutoff points for high-sensitivity C-reactive protein (hs-CRP) as a component of metabolic syndrome (MetS) in Japanese have been proposed as 0.40-0.45 mg/L for men and 0.25-0.35 mg/L for women. However, there are some concerns about the reproducibility of hs-CRP. Methods and Results: Reproducibility of hs-CRP as a component of MetS was examined using receiver-operating characteristic (ROC) curves for diagnosing MetS in 1,274 men and 673 women whose serum levels of hs-CRP were measured twice at annual health screening tests. The Spearman's correlation coefficient between baseline hs-CRP and hs-CRP at the next year's test was 0.68 in men and 0.71 in women. The area under the ROC curves of baseline hs-CRP, hs-CRP at the next year's test, and the mean of the 2 hs-CRP tests for diagnosing baseline MetS were 0.71, 0.71, and 0.72, respectively, in men and 0.75, 0.74, and 0.74, respectively, in women. Optimal cutoff points of baseline hs-CRP, hs-CRP at the next year's test, and the mean of 2 tests for diagnosing baseline MetS were all 0.40 mg/L in men and 0.35 mg/L in women. Conclusions: The serum level of hs-CRP was stable enough for use as a measure of the inflammatory component of MetS, and the optimal cutoff point of hs-CRP was 0.40 mg/L for men and 0.35 mg/L for women in a Japanese health-screening population.  (Circ J 2010; 74: 1488 - 1493)