AD (Alzheimer's disease) is a progressive neurodegenerative disorder characterized by the extracellular accumulation of amyloid β-peptide and the intracellular accumulation of tau. Although there is much evidence linking tau to neurodegeneration, the precise mechanism of tau-mediated neurotoxicity remains elusive. The presence of tau-positive pre-tangle neurons lacking neurofibrillary tangles has been reported in AD brain tissue. In order to study this non-fibrillar tau, we generated a novel monoclonal antibody, named TOC1 (tau oligomeric complex 1), which selectively labels tau dimers and oligomers, but does not label filaments. Time-course analysis and antibody labelling indicates that oligomers appear as an early event in AD pathogenesis. Using a squid axoplasm assay, we have demonstrated that aggregated tau inhibits anterograde FAT (fast axonal transport), whereas monomeric tau has no effect. This inhibition requires a small stretch of N-terminal amino acids termed the PAD (phosphatase-activation domain). Using a PAD-specific antibody, TNT1 (tau N-terminal 1), we demonstrate that PAD exposure is increased in diseased neurons and this leads to an increase in FAT inhibition. Antibody co-labelling with the early-AD marker AT8 indicates that, similar to TOC1, TNT1 expression represents an early event in AD pathogenesis. Finally, the effects of the molecular chaperone Hsp70 (heat-shock protein 70) were also investigated within the squid axoplasm assay. We illustrate that Hsp70 preferentially binds to tau oligomers over filaments and prevents anterograde FAT inhibition observed with a mixture of both forms of aggregated tau. Together, these findings support the hypothesis that tau oligomers are the toxic form of tau in neurodegenerative disease.