This study examined functional trajectories of subjects during the transition phase between ambulatory and non-ambulatory Duchenne muscular dystrophy (DMD) to inform clinical trial designs for new therapeutics. Ambulatory, pulmonary, and upper limb function leading up to loss of ambulation (LoA) and non-ambulatory measures following LoA were quantified; time ordering of pulmonary and upper limb milestones relative to LoA were determined; and the 10-second time threshold for 10-meter walk/run (10MWR) as a marker of approaching LOA was explored. Included in this analysis were 51 subjects aged between 7 and 18 years who experienced LoA during follow-up in the PRO-DMD-01 natural history study. Mean age at LoA was 12.7 (7.1–18.6) years. Mean annual rates of decline in forced vital capacity (FVC) <80%-predicted and performance of upper limb (PUL) 1.2 total score were smaller before than after LoA, but not significantly (FVC %-predicted: 5.6% vs. 10.1%, p = 0.21; PUL 1.2 total score: 2.3 vs. 3.8 units, p = 0.20). More than half of patients experienced clinically significant deficits in FVC %-predicted and PUL 1.2 before experiencing LoA. Among subjects with baseline 10MWR >10 s, those with <1 year to LoA had similar mean ages but significantly worse mean ambulatory function at baseline compared to those with ≥1 year to LoA. Enriching DMD clinical trials for patients with declining pulmonary or upper limb function is achievable without restricting enrollment to non-ambulatory patients. The sequencing of LoA and initial deficits in pulmonary and upper limb function varied across patients and highlights the potential for composite outcomes or multi-outcome trial designs to assess disease-modifying therapies more comprehensively.