An essential prerequisite of randomized controlled trials (RCTs) is “external validation,” that is, results obtained in the “real world” closely replicate those from RCTs. In this respect, contemporary administrative data set registries reporting outcomes after carotid endarterectomy and carotid artery stenting (CAS) have recently reported death/stroke rates that exceed the accepted risk thresholds for intervening. The current article evaluates this controversy and offers reasons why this might continue to happen, namely: (1) difference in interventionists’ competence/experience, (2) patient selection, and (3) advances in CAS technology/technique. As CAS is a continuously evolving technique, the results obtained from patients recruited into the landmark RCTs (as early as the late 1990s) do not reflect contemporary practice. Although RCTs are not always the perfect solution, the process of randomization ensures minimization of selection bias. A possible way forward may be the introduction of prospective, randomized, controlled clinical registries.