The association between preoperative smoking cessation and postoperative pulmonary morbidity was studied prospectively in 200 consecutive patients undergoing an elective coronary artery bypass surgical procedure. Detailed respiratory, cardiovascular, and smoking histories were elicited. Preoperative arterial blood gas analyses and bedside spirometry were performed. Urinary cotinine levels were measured to verify smoking histories. During spirometry, severe angina developed in seven patients, who were hence excluded from the study; one patient died of hemorrhage intraoperatively. An observer unaware of patients' preoperative histories assessed the remaining 192 patients throughout the intraoperative and postoperative periods for pulmonary complications. Postoperative pulmonary complications occurred in a third of the current smokers. Patients who had stopped smoking for 2 months or less had a pulmonary complication rate almost 4 times that of patients who had stopped for more than 2 months (57.1% versus 14.5%). Patients who had stopped smoking for more than 6 months had rates similar to those who had never smoked (11.1% and 11.9%, respectively). Preoperative pulmonary dysfunction, increased pack-years of smoking, prolonged surgical time, and the use of enflurane were independently associated with postoperative pulmonary morbidity (P less than 0.05). We concluded that smoking cessation should occur at least 2 months preoperatively to maximize the reduction of postoperative respiratory complications.