Comparisons of results from different types of surveys are necessary to establish the best and cheapest methods of assessing drug use. This paper reports results from a school survey and a household survey conducted in Ontario. Both surveys were done in 1983 and used unbiased samples of the population in the same age groups. There are some methodological differences between the two studies. Younger students were more likely to report the use of tobacco, alcohol and cannabis in the student survey than in the household survey. Similarly, older students (18 and 19) reported more alcohol use in the student survey. In general, the results indicate that respondents are more likely to give socially desirable answers to questions answered at home than at school. Also, alcohol and drug users may be more likely to be missed in household than in school studies as the former have a much higher non-response rate.