This interlaboratory study evaluated the reproducibility of the assessments of neuritic plaques and neurofibrillary tangles (NFTs)--the hallmark lesions of Alzheimer disease--and compared the staining between the BrainNet Europe centers. To reduce the topography-related inconsistencies in assessments, we used a 2-mm tissue microarray (TMA) technique. The TMA block included 42 core samples taken from 21 paraffin blocks. The assessments were done on Bielschowsky and Gallyas silver stains using an immunohistochemical (IHC) method with antibodies directed to beta-amyloid (IHC/Abeta) and hyperphosphorylated tau (IHC/HPtau). The staining quality and the assessments differed between the participants, being most diverse with Bielschowsky (good/acceptable stain in 53% of centers) followed by Gallyas (good/acceptable stain in 57%) and IHC/Abeta (good/acceptable stain in 71%). The most uniform staining quality and assessment was obtained with the IHC/HPtau method (good/acceptable stain in 94% of centers). The neuropathologic diagnostic protocol (Consortium to Establish a Registry for Alzheimer Disease, Braak and Braak, and the National Institute of Aging and Reagan [NIA-Reagan] Institute) that was used significantly influenced the agreement, being highest with NIA-Reagan (54%) recommendations. This agreement was improved by visualization of NFTs using the IHC/HPtau method. Therefore, the IHC/HPtau methodology to visualize NFTs and neuropil threads should be considered as a method of choice in a future diagnostic protocol for Alzheimer disease.