The purpose of this study was to investigate whether family smoking,rnfriends'smoking, and students' coping styles have significant contributionrnon the prediction of cigarette smoking among university students in mainrncampus of Addis Ababa University. A self-reported questionnairerncomprising students' family and friends smoking status items, copingrnscale, and cigarette use inventory was administered to a random samplernof 658 male students. Chi-square and binary logistic regression analysesrnwere used for analysis. Results obtained through chi-square analysisrndisplayed that fathers' smoking, brothers' smoking and friends' smokingrnassociated significantly with cigarette smoking of students. However, therernwas no association between smokers and non-smokers related tornmothers', sisters' and other family members' smoking. Coping stylesrn(productive, non-productive and reference to others) also associatedrnsignificantly with cigarette smoking of students. Logistic regressionrnanalysis revealed that fathers' smoking, brothers' smoking and friends'rnsmoking significantly predicted cigarette smoking of students. Studentsrnwho have father, brother{s) or friend(s) who smoke are more likely tornpractice smoking behaviors than those students with non-smoker father,rnbrother{s) or friend(s). Similarly, coping styles (productive, non-productivernand reference to others) significantly predicted cigarette smoking ofrnstudents. Students with lower scores on productive coping style and/orrnreference to others coping style, and/or higher score on the non-productiverncoping style indicated more likely to smoke. The logistic regression resultrnalso revealed that friends' smoking was a dominant independent predictorrnof cigarette smoking of students. These results highlight the importance ofrnconsidering smoking by fathers, brothers and friends, and the personalrncoping styles for the early detection of cigarette smoking.