In an attempt to gather information on procedures, criteria, and possible difficulties, we have sent a detailed questionnaire on the postmortem diagnosis of Alzheimer disease (AD) to 169 neuropathologists in the German-speaking countries--Germany, Austria, and Switzerland. Of a total of 104 responses, 65 completed questionnaires were evaluated statistically. Fifty-nine percent of all respondents consider the neuropathologic diagnosis of AD a problem. Although 38% consider themselves able to establish the diagnosis without knowledge of clinical data, 29% do not believe this is possible. Thirty-eight percent opine that the presence of neurofibrillary tangles is not a prerequisite for a positive diagnosis, and 3% believe that examination of the hippocampal formation is unnecessary. For 6% of the respondents, neither a silver stain a silver stain nor thioflavin S is necessary. Ninety-seven percent believe that the availability of explicit diagnostic criteria would be helpful; by contrast, only 43% have exact knowledge of the National Institute of Aging criteria, and only 29% have exact knowledge of the Consortium to Establish a Registry for Alzheimer's Disease (CERAD) criteria. Only every fourth neuropathologist uses these criteria regularly, and the validity of current diagnostic criteria is challenged by a considerable number of respondents. Many pathologists consider the quantification of plaques and tangles problematic, are confused by a lacking correlation of the pathology with the clinical picture, and believe that the gray zone between normal aging and full-fledged AD represents a major diagnostic problem. Our survey shows that there currently is no consensus on diagnostic procedures and criteria used in practice. This stresses the need for precise diagnostic guidelines and continuing information of actively practicing neuropathologists to achieve higher interlaboratory reliability of the diagnosis of AD.