The role of the increase in the common carotid artery (CCA) intima-media wall thickness (IMT) in the atherosclerotic process is questionable. This longitudinal study examined the predictive value of CCA-IMT measured at baseline examination (at sites free of plaques) on the occurrence of atherosclerotic plaques in the extracranial carotid arteries during 4 years of follow-up study in a sample of 1010 subjects aged 59 to 71 years. Ultrasound examinations were performed at baseline and 2 years and 4 years later. The occurrence of carotid plaques during follow-up was defined as the appearance of >/=1 plaque in previously normal carotid segments and/or the appearance of new plaques in the carotid segments that previously had plaques. Carotid plaque occurrence was observed in 185 subjects (18.3%). Age- and sex- adjusted odds ratios of carotid plaque occurrence were 2.66 (95% CI 1.58 to 4.46, P<0.001) in subjects having intermediate baseline CCA-IMT values (quartiles 2 and 3) and 3.67 (CI 2.09 to 6.44, P<0.001) in those having the highest baseline CCA-IMT values (quartile 4) compared with those having the lowest baseline CCA-IMT values (quartile 1). Multivariate adjustment for major cardiovascular risk factors did not alter the results. These findings were observed for men and women as well as for subjects with and without carotid plaques at baseline. This 4-year longitudinal study shows that CCA-IMT predicts carotid plaque occurrence in a large sample of relatively old subjects. It extends the findings obtained from cross-sectional studies and suggests that increased intima-media thickness might occur in an earlier phase of the atherosclerotic process.