Objective Some studies suggest that pelvic inflammatory disease (PID) is a potential risk factor for ovarian cancer. However, only few studies have investigated the association between PID and risk of borderline ovarian tumors. We conducted a population-based cohort study to investigate the association between PID and risk of borderline ovarian tumors. Methods Using various nationwide Danish registries we identified all women in Denmark during 1978–2012, who were born during 1940–1970 (n = 1,318,925). Of these, 81,263 women were diagnosed with PID in the study period, and 2736 women had a borderline ovarian tumor (1290 serous and 1344 mucinous). Hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between PID and risk of borderline tumors were estimated using Cox regression models with adjustment for potential confounders. Results A history of PID was associated with an increased risk of borderline ovarian tumors (HR = 1.39; 95% CI: 1.19–1.61). However, histotype-specific analyses revealed significant variation in risk as PID was only associated with an increased risk of serous borderline tumors (HR = 1.85; 95% CI: 1.52–2.24), but not with mucinous borderline tumors (HR = 1.06; 95% CI: 0.83–1.35). Conclusions PID is associated with an increased risk of serous borderline tumors. Further research on the potential underlying biological mechanisms and on the identification of the subset of women with PID who are at increased risk of serous borderline tumors is warranted.