Knowledge of prognostic factors for pulmonary outcomes in DMD serves a general understanding of natural history and can help facilitate externally controlled studies and the long-term evaluation of novel therapies. We assessed prognostic factors for time to forced vital capacity (FVC) %-predicted < 80% (among n=368 boys free of this outcome at first assessment, 121 of whom eventually had the outcome) and time to FVC %-predicted < 50% (among n=488, 51 of whom eventually had the outcome) using Cox proportional hazards analyses. Data originated from three natural history databases (UZ Leuven, PRO-DMD-01 data provided by CureDuchenne, and Cincinnati Children's Hospital Medical Center) Average follow-up time was ~3 years. Being ambulatory at baseline and having higher baseline FVC %-predicted were significant predictors of longer time to pulmonary milestones. Prognostic accuracy was also compared between models based on age, data source, ambulatory status, and FVC %-predicted and models further incorporating steroid type, height, weight, BMI, and timed 10-meter walk/run velocity in ambulatory boys. The additional prognostic factors increased the pseudo-R2, a measure of the proportional of variability explained, from 0.20 to 0.31 for FVC %-predicted < 80% and from 0.16 to 0.19 for FVC %-predicted < 50%. Stratification of boys based on risk score tertiles produced groups with 2-year risks of reaching FVC %-predicted < 80% of 37%, 15%, and 4%, respectively; median times to FVC %-predicted < 80% in these groups were 2.4, 7.6 and 8.5 years, respectively. A prognostic score incorporating ambulatory status, level of 10MWR function for ambulatory boys, and baseline FVC%-predicted identified patients at significantly different levels of near-term risk of pulmonary decline.