Aims: Genome-wide association studies (GWAS) have identified several risk variants for prostate cancer (pCa) mainly in Europeans, which need to be further verified in other racial groups. We selected six previously identified variants as candidates and to define the association with PCa in Northern Han Chinese. Methods: 749 subjects from Beijing and Tianjin in Northern China were included. Six variants (rs10505474, rs7837328, rs4242384, rs7813, rs486907 and rs1058205) were genotyped by high resolution melting (HRM) assays. The individual and cumulative contribution for of the risk of PCa and clinical covariates were analyzed. Results: Among the six candidate variants, onlyrs10505474, and rs7837328, both locating at 8q24 region, were associated with PCa in our population.rs10505474 (A) was associated with PCa ($OR_{recessive}=1.56$, p=0.006); and rs7837328 (A) was associated with PCa ($OR_{dominant}=1.38$, p=0.042/$OR_{recessive}=1.99$, p=0.003). Moreover, we observed a cumulative effects between them ($p_{trend}=2.58{ imes}10^{-5}$). The joint population attributable risk showed the two variants might account for 71.85% of PCa risk. In addition, we found the homozygotes of rs10505474 (A) and rs7837328 (A) were associated with PCa clinical covariants (age at onset, tumor stage, respectively) ($p_{age}=0.046$, $P_{tumorstage}=0.048$). Conclusion: rs10505474 (A) and rs7387328 (A) at 8q24 are associated with PCa and cumulatively confer risk, suggesting the two variations could determine susceptibility to PCa in the Northern Chinese Han population.