This paper compares data on the prevalence of alcohol and other drug use by students in grades 8 and 12 across four sizes of communities. Data from the American Drug and Alcohol Survey (ADAS), administered in approximately 250 communities during the 1992-93 and 1993-94 school years, were analyzed for four community sizes: very small (population under 2,500); nonmetropolitan, not adjacent to a metropolitan area; nonmetropolitan, adjacent; and metropolitan (population of 50,000 or more). Community size comparisons are presented for drug use patterns by gender; perceived availability of substances and alcohol; and drug-related problems with law enforcement, money, school, family relationships, and other social relationships. Substances investigated include alcohol, marijuana, stimulants, cocaine, crack, inhalants, legal stimulants, LSD, heroin, and daily use of cigarettes and smokeless tobacco. Results indicate a lower aggregate level of drug use among youth from very small, rural communities, and a higher level of drug-related problems among metropolitan youth. There was little difference in alcohol use among 12th-graders by community size, but alcohol use caused more problems for rural youth than for others. The data clearly illustrate that even the smallest communities were not immune to substance abuse problems. However, variability among communities in the same size category reflects the myriad factors influencing substance use. An appendix lists 34 styles of drug involvement in the ADAS reporting system. (Contains 10 data tables and figures and 14 references.) (SV)