C-reactive protein (CRP), an acute phase reactant and marker of inflammation, has been shown to be associated with CRP genetic variants and incident hypertension, but it is unclear whether this link is causal. We therefore conducted a prospective, nested case-control study to examine the relationship between single-nucleotide polymorphisms (SNPs) within the CRP gene, circulating CRP levels and the development of hypertension. Plasma CRP levels and the genotypes of eight SNPs were determined in 2000 unrelated Shanghai residents, including 908 hypertensive individuals and 1092 normotensive individuals. Among the 1092 normotensives, 968 subjects were followed up for 2 years, during which 71 developed hypertension. Plasma CRP levels were independently associated with the development of hypertension in the follow-up study (odds ratio per quartile=1.64; 95% confidence interval: 1.18-2.26; P<0.001). The minor alleles of rs1130864 (P<0.001) and rs3093059 (P<0.001) were significantly associated with elevated CRP levels, and the minor alleles of rs1205, rs1800947 and rs2246469 (all P<0.001) were associated with decreased CRP levels. A haplotype-based analysis strengthened the results of single-locus analysis. However, none of the SNPs or haplotypes was significantly associated with blood pressure, incident hypertension or changes between baseline and follow-up blood pressure levels. Taken together, our findings demonstrated that plasma CRP levels were substantially associated with common genetic variants in the CRP gene and could predict the development of hypertension. However, the relationship between genotype and CRP levels was not associated with a change in hypertension risk.