A conspicuous feature shared by Alzheimers disease as well as a variety of highly prevalent, clinically unrelated neurodegenerative disorders is the occurrence of protein aggregates both intra- and extracellularly. Most of these conditions are characterized at autopsy by the presence of such deposits, typically of fibrillar structure and accompanying extensive neuronal cell loss, displaying a selective brain distribution. The recently discovered similarities of a number of these aggregates with a novel type of experimentally induced protein deposit, formed as a general response to discrepancies in protein turnover and designated the “aggresome”, has prompted speculations about the involvement of the ubiquitin-proteasome system in a process fundamental to neurodegeneration. Consistent with this view, protein aggregates have been regarded in a pathogenic connotation, with most aspects of neurologic pathogenesis being largely attributed to their presence in nerve tissues. However, the neurotoxicity of protein aggregates remains ambiguous as direct evidence substantiating it have long remained elusive. A convergence of evidence now support the notion that the actual culprits might comprise the oligomeric, non-fibrillar intermediates that arise early during the aggregation process, termed protofibrils and that the fibrillar end-stage aggregates themselves might actually serve a neuroprotective function. These intermediates ostensibly resolve many puzzling aspects of neurodegeneration and there is evidence that neurotoxicity is one key operational property they may possess. The above attest to the fact that protein aggregation remains a complex issue with a role far more enigmatic than originally thought but nonetheless important for the understanding of the pathological basis of neurodegenerative disorders.