In Alzheimer's disease (AD), senile plaques containing amyloid-beta (Abeta) are associated with neurodegeneration, yet little quantitative data are available concerning the spatiotemporal patterns of neuronal death that result from exposure to Abeta deposits. Furthermore, plaques are accompanied by ferritin-rich cells but no data exist regarding the spatiotemporal expression of ferritin in response to Abeta. The present study has obtained such data after injecting aged Abeta peptide into the parietal cortex of adult rats. Injected deposits of fibrillar Abeta (1 microliter of 1 mM in saline) were cleared within 7 days but did not cause a significant increase in ferritin expression. Counts of dying neurones showed that human Abeta1-40 killed as many neurones as control injections of saline, while human Abeta1-42 and rat Abeta1-40 killed significantly less. We conclude that the fibrillar Abeta in plaques is not likely to be directly responsible for the neurodegeneration and ferritin expression that occurs in AD.