Spinal Muscular Atrophy (SMA) is a childhood form of motor neuron disease that causes a progressive paralysis that, in its most severe form, results in death before two years of age. There is currently no cure or treatment for SMA. SMA is caused by a reduction in levels of Survival Motor Neuron (SMN) protein, which results in the degeneration of lower motor neurons. This degeneration is first observed at the neuromuscular junction (NMJ), where pre-synaptic nerve terminals belonging to the motor neuron become dysfunctional and degenerate during the early stages of disease. Several previous studies have shown that the some populations of motor neurons appear to have a resistance to SMA pathology, while other neighbouring populations are vulnerable. In this study, we attempted to elucidate the cause of this vulnerability spectrum. Initially, we characterised the relative vulnerability of ten different motor unit pools in an established mouse model of severe SMA and attempted to correlate these vulnerabilities with quantified aspects of motor unit morphology. From this study, no significant correlation could be found with any aspect of motor unit morphology examined, suggesting that morphological parameters of motor neurons do no influence their relative susceptibility. We then attempted to identify changes in basal gene expression between protected and vulnerable pools of motor units using microarray analysis. Motor unit pools were labelled using a retrograde tracer injected into muscles that had previously been identified as having highly vulnerable or resistant motor units. Labelled motor neuron cell bodies were then isolated from the spinal cord using laser capture micro-dissection and RNA was extracted for microarray analysis. From this study, we identified several molecular pathways and individual genes whose expression levels compared the gene expression profiles of vulnerable and resistant motor units. Thus, molecular differences between motor neuron pools likely underlie their relative vulnerability to degeneration in SMA. Lastly, we attempted to identify a novel peptide that could be used to label synapses, including neuromuscular junctions, in vivo. This would allow us to non-invasively visualise degenerating NMJs and other synapses in human patients without the need for a biopsy. Such a tool would be extremely valuable in assessing the effectiveness of drug trials, both in human patients and animal models, and may also contribute to earlier diagnosis of motor neuron disorders. To identify a potentially suitable peptide, we used a phage display library and panned for peptides that specifically bound to the outer surface of synapses using synaptosome preparations. From this panning we successfully enriched two peptides, the sequences of which were used to manufacture fluorescently tagged peptides.