Background: We aimed to compare the rate of neurological events in patients with or without cerebral embolic protection (CEP) during transcatheter aortic valve replacement (TAVR). Methods and Results: Data on clinical end points including neurological events ≤30 days post-TAVR were collected for all patients who underwent transfemoral TAVR in 2 academic tertiary care institutions. Patients were matched through propensity scoring, which resulted in 333 pairs of patients with versus without CEP out of a total of 831 consecutive patients. The median age was 81 (76–85) years, and the median logistic EuroScore was 14% (9%–20%). The CEP group experienced less neurological events at 24 hours (1% versus 4%; P =0.035) and at 30 days (3% versus 7%; P =0.029). There were significantly more disabling strokes in unprotected patients at 30 days (1% versus 4%; P =0.039). CEP was associated with significantly fewer neurological events at 24 hours after TAVR (odds ratio, 0.20; 95% CI, 0.06–0.73; P =0.015) by multiple regression analysis, while age and valve type did not contribute significantly. Overall, 67% (2 of 3) in the CEP versus 83% (10 of 12) in the non-CEP cohort experienced neurological events in protected areas (ie, not dependent on the left vertebral artery). Conclusions: The use of filter-based CEP during TAVR was associated with less neurological events, especially in CEP-protected brain territories.