The prevalence of respiratory diseases in smokers and nonsmokers and the incidence of perioperative respiratory events (PREs) were investigated for patients undergoing general anesthesia. The aim was to quantify well-known problems and to identify possible new associations between smoking and PREs. From July 1992 to December 1994, an automatically readable anesthetic record (ARAR) documented risk factors, demographic data, and PREs. PREs were used as defined by the German Society of Anaesthesiology and Intensive Care. Of 26961 subsequent anesthesias in adults, 7122 (26.4%) were performed in smokers with a prevalence of chronic bronchitis of 23.3% (4.8% in nonsmokers). 1573 PREs occurred in 1397 (5.2%) of all anesthetics. 459 events concerned intubation problems and problems in technical airway management. 1114 specific respiratory events (SPREs) like re-intubation, laryngospasm, bronchospasm, aspiration, hypoventilation/hypoxemia and others had a total incidence of 5.5% in smokers and 3.1% in nonsmokers. The relative risk (RR) of SPREs was 1.8 in all smokers, 2.3 in young (16-39 years) smokers, and 6.3 in obese young smokers. The RR of perioperative bronchospasm was 25.7 in young smokers with chronic bronchitis. The authors conclude that the impact of smoking on perioperative respiratory problems should make anesthetists take this widespread preoperative condition seriously, particularly in young adults. The presented method of incident reporting (based on a national classification) could contribute to future research in anesthetic epidemiology.