Persistence of high-risk (HR-) human papillomavirus (HPV) infection of the uterine cervix increases the risk of cervical cancer. Oral HPV infections are among potential covariates of long-term genotype-specific persistent cervical HR-HPV infections. It is not known whether this persistence reflects inability of the host to reject HPV infections in general. A case-control setting was designed to estimate the covariates of long-term persistent cervical HR-HPV infections using multivariate generalized estimating equation (GEE) models. HPV was detected with PCR using GP05+/GP06+-primers and genotyped for 24 HPVs with a Multimetrix-kit. The cases (n=43) included women who had genotype-specific persistent cervical HR-HPV infection for at least 24 months (24M+) and controls were women who tested repeatedly HPV-negative in their cervical samples (n=52). These women represent a sub-cohort of the Finnish Family HPV Study. The cases differed significantly from the HPV-negative controls in several aspects: they were younger, had a longer mean time to incident oral HPV infection (40.7 versus 23.6 months), longer duration of oral HPV persistence (38.4 versus 14.1 months), and longer time to clearance of their oral HPV infection (50.0 versus 28.2 months). In multivariate GEE analysis, the second pregnancy during the follow up was the only independent predictor with significant protective effect against 24M+ persistent cervical HR-HPV infections, OR of 0.15 (95% CI 0.07-0.34). To conclude, long-term persistent cervical HR-HPV infections are associated with a prolonged clearance of oral HR-HPV infections while new pregnancy protects against persistent cervical HR-HPV infections.