The aim of this article is to review our experience in surgical treatment of carotid atherosclerosis using eversion carotid endarterectomy (ECEA) in 5,034 patients, with particular attention to temporal changes in patients' characteristics, diagnostic approach, surgical technique, medical therapy, and outcome in the early (group A, 1991-1997) versus late (group B 1998-2004) period of ECEA. From January 1991 to December 2004, 5,034 primary ECEAs were performed for high-grade carotid stenosis. Patients treated for restenosis after previous carotid surgery were excluded from the analysis. Group A consisted of 1,714 patients who underwent surgery between 1991 and 1997, and group B consisted of 3,320 patients who underwent surgery between 1998 and 2004. Follow-up included routine clinical evaluation and noninvasive surveillance, with duplex scanning at 1 month after surgery, after 6 months, and annually afterward. Only 3% of patients in group A and 0.6% in group B were asymptomatic, with 23% and 47% of them having preoperative stroke, respectively. In group A, angiography was used for the final diagnosis in 78% of patients. In group B, duplex scanning was performed in 82% of patients and angiography in only 18% (p < .001). Clamping time was shorter in the latter group (12.4 +/- 3.1 vs 14.5 +/- 4.1 min, p < .01). Introperative shunting and regional anesthesia were rarely performed in both groups (1.4% vs. 0.4%, p < .01, and 2% vs 0.3%, p < .001). Total and neurologic morbidity was significantly higher in group A than in group B (6.41% +/- 0.47% vs 4.81% +/- 0.53%, p < .001, and 2.14% +/- 0.31% vs 1.23% +/- 0.29%, p < .001, respectively). Total mortality was also higher in group A than in group B (1.92% +/- 0.24% vs 1.36% +/- 0.50%, p < .05), but although there was a trend toward lower neurologic mortality, it did not reach statistical significance (1.04% +/- 0.5% vs 0.57% +/- 0.25%, p = .074). There was a lower rate of nonsignificant restenosis (< 50%) in group B (2% vs 5%, p < .01), but the incidence of restenosis > or = 50% was identical between the groups (5.5% for both). Our data show that ECEA is a reliable surgical technique for the treatment of atherosclerotic carotid disease. Temporal trends in our patients demonstrated a decline in periopertive mortality and morbidity, despite a higher incidence of preoperative stroke.