Objectives This study investigated the relationship between annual hospital volume and the outcomes in carotid endarterectomy and quantified critical volume threshold for this procedure.Data sources PubMed, EMBASE and the Cochrane library were searched for all articles on the volume-outcome relationship in CEA.Review methods Articles were included if they presented data on post-operative mortality and/or stroke rates and annual hospital volume of CEA. The review conformed to the QUOROM statement. The data were meta-analysed and a pooled effect estimate of volume on the stroke and/or mortality rates from CEA quantified, along with the critical volume threshold.Results Twenty-five articles, encompassing 936 436 CEA, were analysed. Significant relationships between mortality rate and stroke rate and annual volume were seen.Overall, the pooled effect estimate was odds ratio 0.78 [95% confidence interval 0.64 -0.92], in favour of surgery at higher volume units, with a critical volume threshold of 79 CEA per annum.Conclusions Significantly lower mortality and stroke rates were achieved at hospitals providing a higher annual hospital volume of CEA. Hospitals wishing to provide CEA should adhere to minimum volume criteria.