Eighteen adolescents (ages 11-15) underwent 6 independent, randomized, double-blind, placebocontrolled trials. In each trial, participants sampled a noncaffeinated and a caffeinated soda (33 mg per 8 oz) in a 2 day crossover, followed by concurrent access to the same 2 sodas the following 2 days. Four of the 18 (22%) participants met a repeatability criterion, and 1 also met a statistical criterion for reliable caffeine self-administration. Across-subjects ratings of depression, drowsiness, and fatigue were increased on noncaffeinated compared with caffeinated soda sampling days. These results are similar to, but less robust than, previous research in adults showing that 31-36% of coffee drinkers reliably self-administer caffeine and experience adverse effects from caffeine abstinence.Caffeine is the most widely consumed psychoactive drug in adults, and almost all children and adolescents are regular daily or weekly caffeine consumers (Barone & Roberts, 1984). In a large (N = 1,135) nationwide survey of 5-18-year-olds, 98% of them consumed caffeine at least once during the surveyed week (Morgan, Stults, & Zabik, 1982). Their average daily caffeine intake was 37.4 mg or 0.9 mg/kg. The quantities of caffeine consumed on a milligram per kilogram basis were relatively constant across age groups. Tea was the greatest source of caffeine (34% of total intake), followed by carbonated beverages (26%), coffee (22%), and chocolate-containing foods and beverages (17%). A more recent survey of 8-11-yearolds in Massachusetts found lower caffeine intake than previous estimates but a greater amount of intake from carbonated beverages (Ellison, 1993). For all children (N = 78), the average daily caffeine intake was 18.9 mg or 0.6 mg/kg. Carbonated beverages provided 55%, chocolate-containing foods and beverages provided 39%, and tea provided 7% of the total amount of caffeine.The ubiquity of caffeine use across all ages has lead to recent