The aim of the study was to establish 24 month changes in upper limb function using a revised version of the performance of upper limb test (PUL 2.0) in a large cohort of ambulant and non-ambulant boys with Duchenne muscular dystrophy and to identify possible trajectories of progression. Of the 187 patients studied, 87 were ambulant (age range: 7-15.8 years), and 90 non-ambulant (age range: 9.08-24.78). The total scores changed significantly over time (p<0.001). Non-ambulant patients had lower total scores at baseline (mean 19.7) when compared to the ambulant ones (mean 38.4). They also had also a bigger decrease in total scores over 24 months compared to the ambulant boys (4.36 vs 2.07 points). Multivariate model analysis showed that the Performance of Upper Limb changes reflected the entry level and ambulation status, that were independently associated to the slope of Performance of Upper Limb changes. This information will be of help both in clinical practice and at the time of designing clinical trials.