Most research to characterise the molecular consequences of spinal muscular atrophy (SMA) has focused on SMA I. Here, proteomic profiling of skin fibroblasts from severe (SMA I), intermediate (SMA II), and mild (SMA III) patients, alongside age-matched controls, was conducted using SWATH mass spectrometry analysis. Differentially expressed proteomic profiles showed limited overlap across each SMA type, and variability was greatest within SMA II fibroblasts, which was not explained by SMN2 copy number. Despite limited proteomic overlap, enriched canonical pathways common to two of three SMA severities with at least one differentially expressed protein from the third included mTOR signalling, regulation of eIF2 and eIF4 signalling, and protein ubiquitination. Network expression clustering analysis identified protein profiles that may discriminate or correlate with SMA severity. From these clusters, the differential expression of PYGB (SMA I), RAB3B (SMA II), and IMP1 and STAT1 (SMA III) was verified by Western blot. All SMA fibroblasts were transfected with an SMN-enhanced construct, but only RAB3B expression in SMA II fibroblasts demonstrated an SMN-dependent response. The diverse proteomic profiles and pathways identified here pave the way for studies to determine their utility as biomarkers for patient stratification or monitoring treatment efficacy and for the identification of severity-specific treatments.