Background: Duchenne Muscular Dystrophy (DMD) is the most common muscular dystrophy, but it is a rare disease, creating challenges for study design. Randomized trials have shown that steroids improve surrogate endpoints of skeletal muscle function, but many treatment questions lack robust data and necessitate innovative approaches. Methods: DuchenneConnect is a novel patient-report registry and the largest US-based registry for DMD. We compared time from birth to loss of ambulation among three categories of steroid use: current, past and never. To control for possible confounding, we used a Cox proportional hazards model to estimate hazard ratios according to steroid use. Results: Current steroid use is associated with longer time to fulltime wheelchair use in this population (HR 0.32, p < 0.0001). Conclusion: A self-report registry can efficiently amass a large population for long-term follow-up and provides a useful adjunct to randomized trials in the study of a rare genetic disease.