A case-control study of cancer of the larynx was carried out in Kerala, Southern India, on 191 male cancer cases and 549 male hospital-based controls. Risk factors investigated were pan(betel)-tobacco chewing, bidi and cigarette smoking, drinking alcohol and inhaling snuff. Significant positive associations with risk were observed for bidi smoking (p less than 0.001), bidi and cigarette smoking (p less than 0.001) and drinking alcohol (p less than 0.001). A predisposing effect of smoking cigarettes alone approached significance (0.1 less than p less than 0.05). What appeared to be an almost significant protective effect of pan-tobacco chewing (0.1 less than p less than 0.05) was found to be an artefact of confounding with smoking, and indeed a significant predisposing effect was observed of occasional chewing (p less than 0.001). After a stepwise logistic regression to eliminate those factors which were not significant when adjusted for other factors, significant effects remained of durations of bidi smoking and cigarette smoking, daily frequency of bidi and cigarette smoking and duration of alcohol drinking. Relative risks of 7.12, 5.18 and 2.58 were observed for durations of more than 20 years of bidi smoking, cigarette smoking and drinking alcohol respectively, and a relative risk of 12.68 was observed for those smoking more than 20 bidis/cigarettes per day, in each case relative to a baseline of those negative for the relevant habit.