Nicotine is recognized to be the major inducer of tobacco dependence. The smoking of cigarettes as an advantageous delivery system for nicotine, accelerates and aggravates cardiovascular disease, and is causally associated with increased risks for chronic obstructive lung disease, cancer of the lung and of the upper aerodigestive system, and cancer of the pancreas, renal pelvis, and urinary bladder. It is also associated with cancer of the liver, cancer of the uterine cervix, cancer of the nasal cavity, and myeloid leukemia. In 1950, the first large-scale epidemiological studies documented that cigarette smoking induces lung cancer and described a dose-response relationship between number of cigarettes smoked and the risk for developing lung cancer. In the following decades these observations were not only confirmed by several hundreds of prospective and case-control studies but the plausibility of this causal association was also supported by bioassays and by the identification of carcinogens in cigarette smoke. Whole smoke induces lung tumors in mice and tumors in the upper respiratory tract of hamsters. The particulate matter of the smoke elicits benign and malignant tumors on the skin of mice and rabbits, sarcoma in the connective tissue of rats, and carcinoma in the lungs of rats upon intratracheal instillation. More than 50 carcinogens have been identified, including the following classes of compounds: polynuclear aromatic hydrocarbons (PAH), aromatic amines, and Nnitrosamines. Among the latter, the tobacco-specific N-nitrosamines (TSNA) have been shown to be of special significance. Since 1950, the makeup of cigarettes and the composition of cigarette smoke have gradually changed. In the United States, the sales-weighted average "tar" and nicotine yields have declined from a high of 38 mg "tar" and 2.7 mg nicotine in 1954 to 12 mg and 0.95 mg in 1992, respectively. In the United Kingdom, the decline was from about 32 mg "tar" and 2.2 mg nicotine to less than 12 mg "tar" and 1.0 mg nicotine per cigarette. During the same time, other smoke constituents changed correspondingly. These reductions of smoke yields were primarily achieved by the introduction of filter tips, with and without perforation, selection of tobacco types and varieties, utilization of highly porous cigarette paper, and incorporation into the tobacco blend of reconstituted tobacco, opened and cut ribs, and "expanded tobacco." In most countries where tobacco blends with air-cured (burley) tobacco are used, the nitrate content of the cigarette tobacco increased. In the United States nitrate levels in cigarette tobacco rose from 0.3-0.5% to 0.6-1.35%, thereby enhancing the combustion of the tobacco. More complete combustion decreases the carcinogenic PAH, yet the increased generation of nitrogen oxides enhances the formation of the carcinogenic N-nitrosamines, especially the TSNA in the smoke. However, all analytical measures of the smoke components have been established on the basis of standardized machine smoking conditions, such as those introduced by the Federal Trade Commission, that call for 1 puff to be taken once a minute over a 2-s period with a volume of 35 ml. These smoking parameters may have simulated the way in which people used to smoke the high-yield cigarettes; however, they no longer reflect the parameters applicable to contemporary smokers, and especially not those applicable to the smoking of low- and ultra-low-yield filter cigarettes. Recent smoking assays have demonstrated that most smokers of cigarettes with low nicotine yield take between 2 and 4 puffs per minute with volumes up to 55 ml to satisfy their demands for nicotine. The overview also discusses further needs for reducing the toxicity and carcinogenicity of cigarette smoke. From a public health perspective, nicotine in the smoke needs to be lowered to a level at which there is no induction of dependence on tobacco.