Self-reports on alcohol use collected via school-based questionnaires, telephone surveys, and household interviews are central measures in many studies in the alcohol field. The validity of such self-reports remains an issue. Use of biological pipeline procedures is one way in which the quality of self-reports might be improved. The current study tested the effectiveness of a saliva test pipeline procedure in increasing drinking disclosure rates among adolescents in the sixth and eighth grades. Two sixth-grade classes from each of 14 elementary schools (n = 828) and four eighth-grade classes from each of 8 middle schools (n = 754) were selected. Half of the classes in each school were assigned to the pipeline condition and half to the control condition. Each student in the pipeline condition was asked to provide a saliva sample via dental roll before completing a questionnaire that all students (pipeline and control) received. Pipeline students were told that "some of the saliva we collect today will be tested in a laboratory and will provide a biological measure of alcohol use." Sixth- and eighth-grade students exposed to the alcohol procedure reported 5 to 7% higher alcohol use prevalences than students in the control group. While the pattern of improved reporting under the pipeline condition held across four alcohol-use measures and two grade levels, the effect was statistically significant for only one measure. The pipeline procedures used here had small effects on adolescent self-reported alcohol use.