Introduction: Solid organ transplant recipients (SOTRs) are at much higher risk of developing squamous cell carcinoma (SCC) and basal cell carcinoma (BCC), compared to the general population. Previous studies have derived genetics-based predictors (polygenic risk scores, PRS) of SCC and BCC risk in SOTRs by assuming that genetic risk variants act in the same way in the general population as in SOTRs, but this assumption has not been fully tested. Objective: To investigate whether known genetic risk variants for SCC and BCC have different effect sizes in SOTRs versus in non-transplantees, and if a re-weighted PRS would improve risk prediction. Methods: We conducted genome-wide association studies for SCC and BCC separately in the non-transplant general population and in SOTRs, and compared the risks associated with selected common genetic variants for KC risk in SOTR vs non-transplant individuals from the UK Biobank. For regions with an increased log odds ratio in SOTRs, PRSs including these weights were validated in the QSkin study, and applied to the Australian STAR SOTR cohort. Results: Effect sizes for functional variants in MC1R (rs1805007), ASIP (rs6059655), and IRF4 (rs12203592) were much more strongly associated with the risk of KC in SOTRs than in non-transplantees. The proportional increase in the effect sizes ranged from 1.9-fold for rs6059655 and BCC risk (SOTRs log (OR)=0.49, 95%CI=0.00-0.98 vs log (OR)=0.26, 95%CI=0.24-0.30 in non-transplantees) to as high as 4.8-fold for rs1805007 and SCC risk (SOTR log (OR)=0.88, 95% CI=0.41-1.35 vs log (OR)=0.18, 95% CI=0.12-0.24 in non-transplantees). PRS with SOTR derived weights for these SNPs showed improved SCC/BCC risk stratification in the STAR Cohort, with the optimised PRS reclassifying 19% of SCC cases vs 8% using the standard PRS, and 18% of BCC cases vs 12% using the standard PRS. Conclusion: Effect sizes for SCC and BCC risk for genetic variants in the MC1R, ASIP and IRF4 genes are elevated in SOTRs, and correctly weighting these variants improves risk stratification based on polygenic risk.