Considerable geographic variation exists in the relative use of hemodialysis (HD) vs peritoneal dialysis (PD). Studies comparing survival between these modalities have yielded conflicting results. Our aim was to compare the survival of Dutch HD and PD patients. We developed Cox regression models using 16 643 patients from the Dutch End-Stage Renal Disease Registry (RENINE) adjusting for age, gender, primary renal disease, center of dialysis, year of start of renal replacement therapy, and included several interaction terms. We assumed definite treatment assignment at day 91 and performed an intention-to-treat analysis, censoring for transplantation. To account for time dependency, we stratified the analysis into three time periods, >3-6, >6-15, and >15 months. For the first period, the mortality hazard ratio (HR) of PD compared with HD patients was 0.26 (95% confidence interval (CI) 0.17-0.41) for 40-year-old non-diabetics, which increased with age and presence of diabetes to 0.95 (95% CI 0.64-1.39) for 70-year-old patients with diabetes as primary renal disease. The HRs of the second period were generally higher. After 15 months, the HR was 0.86 (95% CI 0.74-1.00) for 40-year-old non-diabetics and 1.42 (95% CI 1.23-1.65) for 70-year-old patients with diabetes as primary renal disease. We conclude that the survival advantage for Dutch PD compared with HD patients decreases over time, with age and in the presence of diabetes as primary disease.