AIMS To determine whether the cause of death could be accurately predicted without the need for a necropsy, and thus to consider whether a "view and grant" system of issuing a cause of death could be introduced into England and Wales. METHOD A one year prospective necropsy study was performed incorporating 568 deaths. Before necropsy, in each case the cause of death was predicted from the available history without examination of the body, and this cause was then compared with the cause of death found at necropsy. RESULTS The ability of the pathologist involved in the study to predict a cause of death before necropsy, either while in the mortuary or as a paper exercise, was shown to vary between 61% and 74% of cases. After the necropsy, the number of correct predicted causes of death ranged from 39% to 46%. Ischaemic heart disease was found to be the most common and most accurately predicted cause of death. Some natural diseases were frequently misdiagnosed, whereas certain types of unnatural disease were always identified correctly. CONCLUSIONS This study highlights the advantages and disadvantages of a view and grant system. Although it identifies a potential use of such a system, in some cases such as natural cardiac disease, because of the potentially high diagnostic error rate, the continuation of the present system of postmortem examination as part of the coroner's enquiry is recommended.