Abstract Background The cost-effectiveness of human papillomavirus (HPV) vaccination in women pre-sexual debut has been demonstrated in many countries. This study aimed to estimate the cost-effectiveness of a 3-dose bivalent HPV vaccination at ages 12 to 55 year in both rural and urban settings in China. Methods The Markov cohort model simulated the natural history of HPV infection and included the effect of screening and HPV vaccination over the lifetime of a 100,000 female cohort. Transition probabilities and utilities were obtained from published literature. Cost data were estimated by Delphi panel using healthcare payers’ perspective. Vaccine cost was assumed Hong Kong listed price. Vaccine efficacy (VE) was based on the PATRICIA trial data assuming VE irrespective of HPV type at all ages on incident HPV. Costs and outcomes were discounted at 3 %. Cervical cancer cases and incremental cost-effectiveness ratio (ICER) for vaccination and screening compared with screening alone were estimated for each vaccination age. Reduced VE in women post-sexual debut were investigated in scenario analyses. Results With 70 % vaccination coverage, a reduction of cancer cases varying from 585 to 33 in rural and 691 to 32 in urban were estimated at ages 12 to 55, respectively. The discounted ICERs of vaccination at any age under 23 years in rural and any age under 25 years in urban were lower than the current threshold. Scenario analyses with lower VE post-sexual debut confirmed the results with age 20 in rural and 21 in urban had consistent lower ICERs. The more ‘catch-up’ cohorts vaccinated at the start of a program, the more cancer lesions are avoided in the long-term. Conclusions Vaccination at any age under 23 years old in rural and any age under 25 years old in urban were cost-effective. Catch-up to the age of 25 years in rural and urban could still be cost-effective.