It has been suggested that the pathological lesions of Alzheimer's disease (AD) spread along neuronal connections. This study was designed to examine this hypothesis in the alvear and perforant pathways, two well-defined neuroanatomical pathways that project from the entorhinal cortex to the hippocampus. Paraffin-sections of hippocampal-entorhinal cortex from 25 AD cases were immunolabelled for tau, beta-amyloid (Abeta) and beta-amyloid precursor protein (betaAPP). We used image-analysis to quantify immunolabelling at both ends of the alvear and perforant pathways. At the beginning and the end of the alvear pathway, area of immunolabelling in microm2 per area of field (72000 microm2) were as follows: tau 349 and 821 (P<0.01), Abeta 349 and 61 (P<0.05) and betaAPP 18 and 73 (P<0.01). Corresponding values for the perforant pathway were tau 421 and 387, Abeta 382 and 115 (P<0.05) and betaAPP 55 and 83. Tau was significantly greater at the end than at the beginning of the alvear pathway, but similar at both ends of the perforant pathway. There was significantly more Abeta at the beginning than at the end of the alvear and perforant pathway. These results at least in part reinforce previous work [19] that tau-rich areas may be neuronally connected to Abeta-rich areas.