This paper provides a summary of a national survey of laboratory testing agencies in Canada that do alcohol and other drug analyses on victims of motor vehicle accidents. The purpose was to examine practices and procedures that would affect the reliability and comparability of information relating to alcohol and drug involvement in traffic fatalities, since such data are used extensively for policy making, information programmes, and educational purposes. Technical portions of responses to the survey were evaluated by a panel of toxicologists nominated by the Canadian Society of Forensic Science. Results showed there is considerable uniformity in the practices of laboratories, but substantial discrepancies exist in many areas. Comments on the practices are given together with recommendations for consistency.