Background. To study effects of ever used cigarettes or cannabis on amount and consequences of alcohol drinking. The ATLAS Project is a long-term longitudinal study comparing the development of substance use from high school to the young adult life period (18-23 years) in the US and Sweden. Methods. Baseline data for 3352 17-19 year-old students (65% from Sweden, 35% from US). Measurements: Questionnaire AUDIT-C, Rutgers Alcohol Problem Index, RAPI, ever use of marijuana, cigarettes and alcohol, onset of alcohol drinking, Conduct problems, SCL-8, Impulsivity. Results. The patterns of ever use differed distinctly between the two countries. In US 35% had never used alcohol, 17% had used alcohol and cannabis, 4% alcohol and cigarettes and 22% alcohol, cannabis and cigarettes. Corresponding figures in Sweden were 7%, 1%, 48% and 17%, respectively. Sweden had higher scores on AUDIT C and RAPI than US in the first three groups. In hierarchical multiple regression analyses on AUDIT C early drinking onset, conduct problems and impulsivity had about the same positive effects. Sweden, ever use of cigarettes and marijuana were all significantly related. Marijuana influenced less in Sweden than in the US. In hierarchical multiple regression analyses on RAPI conduct problems had a very strong effect. Sweden, cigarettes and marijuana were all significantly related. Cigarettes influenced less on consequences of drinking in Sweden than in the US. Conclusions. Sweden and the US differed in important aspects on modulating effects of cannabis and cigarettes on alcohol drinking. Supported by a grant from NIAAA/NIH for Larimer/Berglund.