Abstract Background We previously investigated the association between five ‘first-generation’ measures of epigenetic aging and cancer risk in the Melbourne Collaborative Cohort Study. The present study assessed cancer risk associations for three recently developed methylation-based biomarkers of aging: PhenoAge, GrimAge, and predicted telomere length. Methods We estimated rate ratios (RRs) for the association between these three age-adjusted measures and risk of colorectal (N = 813), gastric (N = 165), kidney (N = 139), lung (N = 327), mature B-cell (N = 423), prostate (N = 846) and urothelial (N = 404) cancer, using conditional logistic regression models. We also assessed associations by time since blood draw and by cancer subtype, and investigated potential non-linearity. Results We observed relatively strong associations of age-adjusted PhenoAge with risk of colorectal, kidney, lung, mature B-cell, and urothelial cancers (RR per standard deviation [SD]was approximately 1.2–1.3). Similar findings were obtained for age-adjusted GrimAge, but the association with lung cancer risk was much larger after adjustment for smoking status, pack-years, starting age, time since quitting and other cancer risk factors (RR per SD = 1.82, 95%CI = 1.44–2.30). Most associations appeared linear, larger than for the first-generation measures, and were virtually unchanged after adjustment for a large set of sociodemographic, lifestyle and anthropometric variables. For cancer overall, the comprehensively-adjusted RR per SD was 1.13 (95%CI = 1.07–1.19) for PhenoAge and 1.12 (95%CI = 1.05–1.20) for GrimAge, and appeared larger within 5 years of blood draw (RR = 1.29, 95%CI = 1.15–1.44 and 1.19, 95%CI = 1.06–1.33, respectively). Conclusion The methylation-based measures PhenoAge and GrimAge may provide insights into the relationship between biological aging and cancer and be useful to predict cancer risk, particularly for lung cancer.