Objective. To measure the long-term rate of radiographic progression in a cohort of patients treated early vs late with conventional DMARDs. Methods. The long-term rate of radiographic progression in patients included in the Swiss clinical quality management in rheumatoid arthritis (SCQM-RA) registry who initiated treatment with conventional DMARDs within the first year of symptom onset (early DMARD) vs patients who initiated treatment 1-5 years after symptom onset (late DMARD). Radiographic progression was assessed in 38 joints using a validated score (Ratingen Score). The rate of progression was calculated using a multivariate regression model for longitudinal data, adjusting for potential confounders. Results. A total of 970 RA patients were included. The 368 patients in the early DMARD group started therapy after a median symptom duration of 6 months, whereas the 602 patients in the late DMARD group initiated therapy after median 2.5 years. RF, MTX use and other risk factors for erosive disease progression were similar between the two groups. However, the estimated rate of radiographic progression at baseline was higher in the early DMARD vs the late DMARD group (1.8 vs 0.6, P < 0.01). In spite of this, the long-term rate of radiographic progression was significantly lower in the early DMARD group after adjustment for confounding factors (-0.35 at 5 years, P = 0.012). Conclusion. This result supports the concept of a therapeutic window of opportunity early in the disease course and suggests that early initiation of DMARD therapy results in a long-lasting reduction of radiographic damage.