School surveys have become the most common method for determining the general level of tobacco, alcohol and illicit drug use among adolescents in a large number of countries. The two most commonly applied field procedures in school surveys are teachers' and researchers' administration of questionnaires. The impact of the difference between these two modes of administration has not been determined, although it has been argued that central involvement of teachers in data collection may decrease the perceived anonymity of the survey among students, and should thus be avoided. In a split-half random sample of 3017 urban, Icelandic 16-20-year-old students, no statistically significant differences were found in reported use of licit or illicit drugs, nor in the stated willingness to be honest between these two modes of administration. These results suggest that when students return their responses in sealed envelopes, the two modes of administration produce equally valid results.