Introduction: The deleterious health effects of tobacco use and smoking in particular have been well documented. Anti-smoking campaigns have been in existence for decades. Doctors, generally perceived as role models by members of their societies, can play a significant role in anti-tobacco programmes. Objectives: The objectives were: to determine the prevalence of tobacco use among medical students, their knowledge of the health effects of tobacco use, their beliefs and perceptions about anti-tobacco legislation/restrictions and the role of doctors as advocates in tobacco-cessation programmes. Methodology: A cross-sectional census was carried out using the 3rd and 5th year students at the medical school of the University of the Witwatersrand (n = 357). Results: The total prevalence of smoking was 15.4% (n = 53), with 9.6% (n = 34) smoking occasionally and 6.2% (n = 22) smoking daily. Among the smokers, more of the 3rd year students smoked daily compared to the 5th years (64% vs 19.3%, p = 0.001). Overall, there were no major differences between the sexes with respect to prevalence of smoking (P= 0.312 Fischer exact test) across both classes. Within the male population (Africans, whites and Asians) overall differences ( 2- אל 11.1, p= 0.006) existed in their smoking patterns. Both the whites and the Asians smoked more than their African counterparts (5.8%, n = 3). There were however no differences in the smoking prevalence rates between the whites and the Asians ((20.9% vs 31.6%: p= 0.314). The females differed (P = 0.042 Fischer exact test) in their smoking behaviour across the races: Coloured (33.3%, n = 2), white (16.5%, n = 18), Asian (13.3%, n =9) and Africans (4.0%, n = 2). The prevalence of noncigarette tobacco use was 14%, with the majority smoking hookahs (7.1%, n=26). More of the 5th year students than their 3rd year students counterparts (23.5% vs 7.1%) had very good knowledge of tobacco cessation techniques and overall differences were observed ( 2אל -35.5, P <0.001). About 90% (n = 316) of the students knew about the dangers of smoking, but 12% (n = 41) did not associate smoking with heart disease. About 45% (n= 179) of the respondents believed that ban of smoking in public space had reduced prevalence rates and about 40% (n= 172) believed that taxation had reduced prevalence rates of tobacco use in South Africa. Almost 50% (n= 164) students believed that the ban of tobacco sales to minors had reduced prevalence rates and about 85% (n= 286) of the respondents agreed that pictorial health warnings should be placed on the packages of tobacco products. The students generally believed that doctors are seen as role models (84%, n= 286) by society and should not smoke (86.3%, n= 299). Even though 81% (n= 279) of the students were willing to speak to the community about the dangers of tobacco, only about 30% (n= 101) were willing to be members in an anti-tobacco organisation. Discussion and conclusion: The prevalence of smoking among the respondents was lower than national prevalence rates among adults but similar to that of South African young adults. The students had a low knowledge of tobacco cessation techniques. Less than half of the respondents believed that the current anti-tobacco legislation was effective and the majority supported the inclusion of pictorial warnings on the packages of tobacco products. The majority of the participants agreed that doctors were perceived as role models by members of their communities and should not smoke. There was a general willingness on the part of the respondents to advice their patients to quit smoking but majority of the participants however would not want to get involved in anti-tobacco organisations in the future.