Background and objectiveThe European Lacidipine Study of Atherosclerosis (ELSA) is a prospective, randomized, double-blind, multi-national interventional trial to determine the effect of four-year treatment using the calcium antagonist lacidipine versus the β-blocker atenolol on the progression of carotid atherosclerosis in 2259 asymptomatic hypertensive patients. B-mode ultrasound is used to measure the primary and secondary endpoints including the mean maximum intima-media thickness (IMT) of the carotid bifurcations and the common carotid arteries (CBMmax), the mean maximum IMT of 12 standard carotid sites (Mmax) and the overall maximum IMT (Tmax). This paper reports the cross-sectional reproducibility of ultrasound measurements at baseline.MethodTo evaluate measurement reliability, each patient is scanned twice at baseline and again at four annual visits, with 80% of the replicate scans performed by the same sonographer and 20% by a different sonographer; 50% of the replicate scans are read by the same reader and the other 50% by different readers.ResultsThe overall coefficient of reliability (R) was 0.859 for CBMmax, 0.872 for Mmaxand 0.794 for Tmax. The reliability for CBMmaxwas stable during the 13/4-year baseline period (R= 0.848 to 0.953) and was uniform among the 23 field centres (R= 0.798 to 0.926). Intra-and inter-reader reliability were 0.915 and 0.872 respectively, and intra-sonographer reliability was 0.866.ConclusionThe results demonstrate that by implementing standardized protocols and strict quality control procedures, highly reliable ultrasonic measurements of carotid artery IMT can be achieved in large multi-national trials.