A consecutive series of 233 carotid endarterectomies performed over a 4-year period by one neurosurgeon was reviewed to analyze the number, nature, cause, and time to occurrence of complications. The time to discharge was analyzed regarding the type of anesthesia used and the time period during which surgery was performed. The records were reviewed to determine whether any complication could have been prevented by routine postoperative monitoring in an intensive care unit or by longer postoperative hospitalization. There were no perioperative deaths. The incidence of ipsilateral stroke was 1.7% and of major stroke was 0.9%. Nondisabling myocardial infarctions occurred in 1.3% of the patients. Complications occurred in a bimodal time course, less than 48 hours or greater than 1 week. The time to discharge decreased during the study period from 4.44 +/- 4.22 days for the first year of the study to 1.56 +/- 1.31 days for the last year of the study (P < 0.0001). A trend toward earlier discharge for patients who received cervical block anesthesia rather than general anesthesia did not reach statistical significance. No adverse patient outcome could be attributed to lack of intensive care unit monitoring or to early discharge. Our data demonstrate that patients who have undergone carotid endarterectomies do not routinely require intensive care unit monitoring and that discharge 24 to 48 hours after surgery is safe for the majority of patients. This regimen can result in considerable savings without compromising the quality of patient care.