Results showed that there was already a notable prevalence of substance use in the first graders, which was slightly higher in the fourth graders, and which had increased more between the fourth and seventh graders. The differences in consumption between the three groups may reflect cohort or period effects, but considering the small age differences between the groups, they are more likely to reflect aging effects. Beer, wine, hard liquor, marijuana, and cigarette smoking all increased with age; the only exception was glue sniffing, which had a rather low and stable prevalence across the different grades. Otherwise, the age trends were evident from both the lifetime and the 6-month prevalence rates. The differences in prevalence of substance use between the three groups may have been even more pronounced if "nonsecret" use of substances had been included in the reporting of substance use for the first and the fourth graders. 4 tables and 26 references (Author abstract modified)