OBJECTIVE: To compare outcomes by steroid treatment among non-ambulatory DMD patients. BACKGROUND: Corticosteroids are the standard of care for DMD; however, steroid use varies after loss of ambulation. Additional evidence on the clinical impact of different steroids is needed in this population. DESIGN/METHODS: A cohort of non-ambulatory DMD patients was identified from PRO-DMD-01: a prospective, observational study of DMD disease progression. Associations between steroid treatment (prednisone, deflazacort, or no steroids) and pulmonary, cardiac, and functional outcomes were assessed, including changes in forced vital capacity [FVC] %-predicted, left ventricular ejection fraction [LVEF], performance of upper limb [PUL] score, and loss of hand-to-mouth function. Outcomes were assessed using Kaplan-Meier analyses and Cox proportional hazards models for milestones, and mixed models with repeated measures for longitudinal outcomes. Models adjusted for selected baseline characteristics (e.g., age, steroid duration). RESULTS: 86 non-ambulatory patients (mean age 13.4 years; n=40 deflazacort; n=29 prednisone; n=17 no steroids) were included. Relative to no steroids, both steroids were associated with delays in median age at FVC %-predicted<60% (+0.9 [prednisone]; +2.3 [deflazacort]; log-rank p<0.01). Median ages at LVEF<55% were numerically prolonged, but non-significant (+2.7 [prednisone]; +0.8 [deflazacort]; p=0.65). While median ages at loss of hand-to-mouth function were not consistently reached, higher proportions of steroid patients maintained function at age 15 (85%-deflazacort; 83%-prednisone; 78%-no steroids; p<0.001). In adjusted Cox models, both steroids showed a significant delay in all three milestones relative to no steroids. In longitudinal models for change in PUL, prednisone patients had significantly slower decline compared to no steroids (+2.5 points/year; p=0.03), and deflazacort patients were significantly slower than prednisone (+1.5 points/year; p=0.04). Changes in FVC %-predicted and LVEF indicated significantly slower decline for both steroids relative to none. CONCLUSION: Steroid use after loss of ambulation was associated with delayed progression of important pulmonary, cardiac and functional deficits in DMD.