Reports the results of a detailed survey of practices followed by 62 cancer registries located throughout the European Union. Concerned with the needs of researchers as well as cancer registries, the study aimed to determine the extent to which different registries follow uniform practices and thus produce internationally comparable data. In particular, the survey responded to the need, in modern epidemiological and clinical research, for data on cancer incidence that are more precise and reliable than mortality statistics. In conducting the survey, IARC also aimed to identify changes in routine data collection and recording that could strengthen the value of registry information for research. With these goals in mind, the study was designed to elicit numerous details about the operational characteristics of Europe's cancer registries. The study's purpose, design, and methods are described in the first section. Section two presents the abundant information collected on types of data recorded, information sources, coding procedures, criteria used to assign the date of incidence, classification of tumor site and stage, the basis of diagnosis, treatment information included follow-up and recording of death, measures to ensure confidentiality, and validity checks for quality control. The results are discussed in section three, which concentrates on the consistencies and variations identified by the study. Variations in the criteria used to record the date of incidence are underscored in view of their implication for survival studies. Of the changes that could be introduced to enhance the comparability of registry data, the report cites compliance with contemporary classification systems for tumor morphology and topography, greater access to death certificate information, and a standardized approach to quality control. The report also identifies several relatively simple changes in registry working practices that could improve the comparability of statistics and greatly enhance their research value. The report concludes with a reproduction of the questionnaire used and a detailed tabulation of data for each registry included in the study. These tables will enable researchers to specify precisely the characteristics of registries to be included in comparative studies while also facilitating the interpretation of differences observed between registries in studies already published.