Smoking and the validity of information obtained on it is often questioned in view of the widespread belief that adolescents tend to under- or over-report the habit. The aim here was to verify smoking habits as reported in a questionnaire given in conjunction with dental examinations by asking participants directly whether they smoked or not and performing biochemical measurements of thiocyanate in the saliva and carbon monoxide in the expired air. The series consisted of 150 pupils in the ninth grade (age 15 years). The reports in the questionnaires seemed to provide a reliable estimate of adolescent smoking, the sensitivity of the method being 81-96%, specificity 77-95%. Biochemical verification or control of smoking proved needless in normal dental practice. Accepting information offered by the patient provides a good starting point for health education and work motivating and supporting of self-directed breaking of the habit.