Introduction Rigorous estimates for the duration of untreated chlamydia infection are important for understanding its epidemiology and designing control interventions, but are only available for women. We have estimated the duration of untreated infection in men.

Methods Data came from published studies in which untreated, chlamydia-infected men were re-tested at a later date. We used analysis methods that had previously been applied to data from women, which allow for a new infection to take one of multiple courses, each clearing at a different rate. We determined the optimal number of possible courses. Parameter estimates were obtained using a Bayesian statistical framework.

Results The best-fitting model had two different courses of infection: ‘slow-’ and ‘fast-clearing’, as had been the case for women. In men only 68% (57%–78%) (median sample; 95% credible interval) of incident infections were ‘slow-clearing’, compared with 77% (69%–84%) in women. The posterior median estimate for the mean infection duration in men was 2.84 (0.87-18.79) years, compared with 1.35 (1.13–1.63 years) in women.

Discussion Our estimated infection duration in men is longer than has previously been assumed. Male infections are less likely to become established (slow-clearing) than those in women but once established, tend to last longer. Long-term, asymptomatic infections in men – in whom chlamydia screening rates are lower – could be sustaining chlamydia prevalence in both sexes. This study provides an improved description of chlamydia’s natural history to better inform public health decision-making. We advocate further data collection to reduce uncertainty in estimates.