We determined the effect of sample size on performance of polygenic hazard score (PHS) models in prostate cancer. Age and genotypes were obtained for 40,861 men from the PRACTICAL consortium. The dataset included 201,590 SNPs per subject, and was split into training and testing sets. Established-SNP models considered 65 SNPs that had been previously associated with prostate cancer. Discovery-SNP models used stepwise selection to identify new SNPs. The performance of each PHS model was calculated for random sizes of the training set. The performance of a representative Established-SNP model was estimated for random sizes of the testing set. Mean HR_98/50 (hazard ratio of top 2% to average in test set) of the Established-SNP model increased from 1.73 [95% CI: 1.69–1.77] to 2.41 [2.40–2.43] when the number of training samples was increased from 1 thousand to 30 thousand. Corresponding HR_98/50 of the Discovery-SNP model increased from 1.05 [0.93–1.18] to 2.19 [2.16–2.23]. HR_98/50 of a representative Established-SNP model using testing set sample sizes of 0.6 thousand and 6 thousand observations were 1.78 [1.70–1.85] and 1.73 [1.71–1.76], respectively. We estimate that a study population of 20 thousand men is required to develop Discovery-SNP PHS models while 10 thousand men should be sufficient for Established-SNP models.