Abstract Introduction Tobacco uptake in adolescents is associated with a range of predictors. We examined the predictors of cigarette smoking, smokeless tobacco (ST) consumption, and use of both ST and cigarettes among adolescents in four South Asian countries. Methods We analyzed the Global Youth Tobacco Surveys (GYTS) data for Bangladesh (2013), India (2009), Pakistan (2013), and Sri Lanka (2015), using multinomial regression to examine associations between several predictors and tobacco use. Results Data from 23 681 adolescents were analyzed. Overall, 82.8% of the study population were between 13 and 15 years and 52.7% were girls, 2% were cigarette smokers, 6.5% were ST users, and 1.1% used both ST and cigarettes, in the past 30 days. Exposure to smoking in public places was associated with past 30-day smoking (relative risk ratio [RRR] 5.59, 95% confidence intervals [CI] 4.28–7.28), ST use (RRR 2.07, 95% CI 1.84–2.32), and use of both ST and cigarettes (RRR 11.42, 95% CI 7.44–17.54). Exposure to tobacco use in electronic media and being offered free tobacco products were associated with all forms of tobacco use. Shopkeepers’ refusal to sell cigarettes protected adolescents from smoking (RRR 0.47, 95% CI 0.36–0.63) and ST use (RRR 0.65, 95% CI 0.45–0.95). However, exposure to antitobacco mass media messages was not protective for any form of tobacco use. Adolescents taught at school about harmful effects of tobacco were less likely to use ST; no evidence of this association was observed for smoking. Conclusion The associations between tobacco use and protobacco factors were strong, but the associations with antitobacco factors lacked strength and consistency in this study population. Implications The predictors of adolescents using different tobacco products, crucial to inform and evaluate tobacco control efforts, are poorly understood. We investigated the associations between several environmental-level factors and cigarette smoking, ST consumption, and use of both forms among adolescents, whereas most of the previous studies focused on individual-level factors. Our study found strong associations between tobacco use and protobacco factors and lack of strength and consistency in associations between antitobacco factors and tobacco use in the study population. Our results indicate that the current tobacco control policies need strengthening to curb the tobacco epidemic in these countries.