Samenvatting
Numbers describing alcohol-involvement in annual trafficdeath tolls, e.g. 27-000 - 35.000, are exaggerated becausestudies have been misinterpreted,misread and sometimesmisrepresented. Every study of alcohol-involvcment in fatalcrashes is biased because of incomplete data which show ahigher proportion of heavy drinkers than would be found ifall victimes could be tested.The proportion of traffic deathscausally related to alcohol may be on the order of 30p. Amethod for learning about alcohol involvement in survivingdrivers is described.