Although previous studies have examined the burden of cancer attributable to tobacco smoking, updated estimates are needed given the dramatic changes in smoking behaviours over the last 20 years. In this study, we estimate the proportion of cancer cases in 2015 attributable to past tobacco smoking and passive exposure in Canada and the proportion of cancers in the future that could be prevented through the implementation of interventions targeted at reducing tobacco use. Data from the Canadian Community Health Survey (2003) were used to estimate the prevalence of active tobacco smoking and passive exposure. Population attributable risk estimates were employed to estimate the proportion of cancers attributable to tobacco in 2015. The prevalence of active tobacco smoking and passive exposure was projected to 2032 and cancer incidence was projected from 2016 to 2042 to estimate the future burden of cancer attributable to tobacco. In 2003, 30% and 24% of Canadians were former and current smoker, respectively and 24% had been exposed to tobacco smoke in the past. We estimated that 17.5% (32,655 cases; 95% CI: 31,253–34,034) of cancers were attributable to active tobacco smoking and 0.8% (1408 cases; 95% CI: 1048–1781) to passive tobacco exposure in never smokers. Between 41,191 and 50,696 cases of cancer could be prevented by 2042 under various prevention scenarios. By decreasing passive tobacco exposure by 10–50%, between 730 and 3650 cancer cases could be prevented by 2042. Strategies focused on reducing the prevalence of tobacco smoking are crucial for cancer control in Canada. •17.5% of cancers diagnosed in Canada in 2015 were attributable to tobacco smoking.•0.8% of cancers diagnosed in Canada in 2015 were attributable to passive tobacco smoke exposure.•Over 70% of lung cancer was attributable to tobacco smoking.•Over 50,000 cases of cancer could be prevented by 2042 by reducing the prevalence of tobacco smoking by 50%.