BACKGROUND: Anticipating the future burden of chronic obstructive pulmonary disease (COPD) is required to develop adequate public health policies. METHODS: A dynamic population model was built to estimate COPD prevalence by 2025 using data collected during the most recent large general population study on COPD prevalence in France (2005) as baseline values. Sensitivity analyses were performed to test the effect of variations in key input variables. RESULTS: The model predicted a steady increase in crude COPD prevalence among subjects aged>/=45 years from 2005 (prevalence estimate: 84.51 per thousand) to 2025 (projected prevalence: 95.76 per thousand, + 0.56 per thousand/yr). There was a 4-fold increase in the prevalence of GOLD grade 3-4 cases, a 23% relative increase in women and a 21% relative increase in subjects >/=75 years. In sensitivity analyses, these temporal trends were robust. Factors associated with > 5% relative variations in projected 2025 prevalence estimates were baseline prevalence and severity distribution, incidence in women and severity of incident cases, transition rates between severity grades, and mortality. CONCLUSIONS: Projections of future COPD epidemiology consistently predict an increase in the prevalence of moderate-to-very severe COPD, especially due to increases among women and subjects aged >/=75 years. Developing robust prediction models requires collecting reliable data on current COPD epidemiology.Read less <