The length of hospital stay after coronary surgery was studied in 4,683 patients undergoing cardiac catheterization followed by coronary surgery at Emory University Hospital or Crawford Long Hospital between the years 1981 and 1986. Length of stay after coronary surgery had a median and modal value of 7 days. There was, however, a long statistical tail of patients with a prolonged length of stay extending out to more than 180 days. Prolonged length of stay (greater than 10 days) could be correlated with preprocedural variables such as age, elective versus emergency status, angina class, ejection fraction, and gender. Length of stay increased from a mean of 6.9 +/- 1.4 days under the age of 40 years to 10.9 +/- 12.1 days over the age of 70 years (p less than 0.0001). Length of stay was correlated with the periprocedural variables of wound infection, neurologic event, arrhythmias, pneumonia, postoperative myocardial infarction, mortality, and pericarditis. Length of stay increased from 8.8 +/- 9.6 days without a neurologic event to 21.1 +/- 17.9 days with a neurologic event (p less than 0.0001). Similarly, without a wound infection, the average stay was 8.7 +/- 8.9 days; with a wound infection, the average stay was 32.2 +/- 25.8 days (p less than 0.0001). The correlates of prolonged stay were tested in another population comprising 781 patients undergoing cardiac catheterization followed by coronary artery bypass grafting in 1987. The predictors of prolonged stay in the 1987 population were wound infection, pneumonia, arrhythmias, age, neurologic events, postoperative infarction, and ejection fraction. Thus, length of hospital stay after coronary surgery may be predicted by multiple preprocedural and periprocedural variables.(ABSTRACT TRUNCATED AT 250 WORDS)