Toxicity associated with abnormal protein folding and protein aggregation are major hypotheses for neurodegeneration. This article comparatively reviews the experimental and human tissue-based evidence for the involvement of such mechanisms in neuronal death associated with the motor system disorders of X-linked spinobulbar muscular atrophy (SBMA; Kennedy's disease) and amyotrophic lateral sclerosis (ALS), especially disease related to mutations in the superoxide dismutase (SOD1) gene. Evidence from transgenic mouse, Drosophila and cell culture models of SBMA, in common with other trinucleotide repeat expansion disorders, show protein aggregation of the mutated androgen receptor, and intraneuronal accumulation of aggregated protein, to be obligate mechanisms. Strong experimental data link these phenomena with downstream biochemical events involving gene transcription pathways (CREB-binding protein) and interactions with protein chaperone systems. Manipulations of these pathways are already established in experimental systems of trinucleotide repeat disorders as potential beneficial targets for therapeutic activity. In contrast, the evidence for the role of protein aggregation in models of SOD1-linked familial ALS is less clear-cut. Several classes of intraneuronal inclusion body have been described, some of which are invariably present. However, the lack of understanding of the biochemical basis of the most frequent inclusion in sporadic ALS, the ubiquitinated inclusion, has hampered research. The toxicity associated with expression of mutant SOD1 has been intensively studied however. Abnormal protein aggregation and folding is the only one of the four major hypotheses for the mechanism of neuronal degeneration in this disorder currently under investigation (the others comprise oxidative stress, axonal transport and cytoskeletal dysfunctions, and glutamatergic excitotoxicity). Whilst hyaline inclusions, which are strongly immunoreactive to SOD1, are linked to degeneration in SOD1 mutant mouse models, the evidence from human tissue is less consistent and convincing. A role for mutant SOD1 aggregation in the mitochondrial dysfunction associated with ALS, and in potentially toxic interactions with heat shock proteins, both leading to apoptosis, are supported by some experimental data. Direct in vitro data on mutant SOD1 show evidence for spontaneous oligomerization, but the role of such oligomers remains to be elucidated, and therapeutic strategies are less well developed for this familial variant of ALS.