The aged canine displays many features that make it an excellent model for studying the progression of pathology in brain aging and linking these findings to learning memory and other cognitive functions. Canines develop extensive β-amyloid deposition within neurons and their synaptic fields which appears to give rise to senile plaques. These plaques are primarily of the early diffuse subtype. Aged canines also exhibit accumulations of lipofuscin cerebral vascular changes, dilation of the ventricles and cytoskeletal changes. Neurofibrillary tangles (NFT s) are not present in the aged canine. Thus, the aged canine brain provides a suitable model for studying early degeneration normally considered to be pre-Alzheimer's. This supposition is also supported by behavioral data. We have found that the extent of β-amyloid deposition correlates with a decline in select measures of cognitive function. These data provide the first evidence of a correlation between β-amyloid accumulation and cognitive decline in the absence of NFTs. We summarize four lines of evidence that support using the aged canine as a model of human aging: (a) Aged canines develop aspects of neuropathology similar to that observed in aged humans; (b) Veterinarians have observed that many canines exhibit a clinical syndrome of age-related cognitive dysfunction; (c) Aged canines are deficient on a variety of neuropsychological tests of cognitive function; (d) The level of β-amyloid accumulation correlates with cognitive dysfunction in the canine. These data indicate that the aged canine is a particularly useful model for studying age-related cognitive dysfunction (ARCD), early neuronal changes associated with aging, and the initial stages of senile plaque formation.