As high-sensitivity C-reactive protein (hsCRP) seems to be associated with an increased risk of cognitive decline, this nested case-control study examined the relation of hsCRP and mild cognitive impairment (MCI) at different time points. 148 MCI cases (106 amnestic, 42 non-amnestic (aMCI/naMCI)) and 148 matched controls were identified from a prospective population based cohort study of 4,359 participants (aged 50-80). HsCRP levels were measured 5 years before (baseline) and at the time of neuropsychological testing (follow-up). Odds ratios (OR) for hsCRP quartiles serum levels were calculated for the two time points using logistic regression analyses and were adjusted for cardiovascular covariates. In the fully adjusted model, baseline hsCRP levels were significantly associated with both MCI and aMCI (OR = 2.29, 95% confidence interval (CI), 1.01-5.15, first versus fourth quartile, respective OR = 2.73, 95% CI, 1.09-6.84). At follow-up, the fourth hsCRP quartile was associated with MCI (OR = 3.60, 95% CI, 1.55-8.33), aMCI (OR = 3.73, 95% CI, 1.52-9.17) and naMCI (OR = 3.66, 95% CI, 1.00-13.77). Elevated hsCRP levels, even detected five years before diagnosis, are associated with an at least twofold increased probability of MCI. These findings suggest that inflammation plays an important role in the development and presence of cognitive impairment.