X chromosome inactivation studies indicate that smooth muscle cells in human atherosclerosis are monoclonal. Monoclonality could arise either by 1) proliferation of a single cell in the adult intima, eg, by selection or mutation, or 2) proliferation of many cells within a large, pre-existing clonal patch that formed during development. To determine whether clonal expansion occurs concomitantly with plaque growth or as part of normal development, X chromosome inactivation patterns were mapped in microdissected samples of aortic smooth muscle, using the human androgen receptor locus as a marker. As expected, 43% of plaque samples were skewed toward one X chromosome, indicating a monoclonal population. Surprisingly, 25% of normal medial samples and 31% of diffuse intimal thickening samples also were skewed toward one X chromosome, indicating a relatively large patch size. Furthermore, 30% of diffuse intimal thickening and 22% of medial samples showed contiguous regions of 4 mm skewed to the same allele, showing that patch length often exceeded 4 mm. Intima and overlying media typically were skewed to the same allele (73% concordance), suggesting common cells of origin. Because patch size is large in normal arteries, X-inactivation analysis cannot discriminate between a monoclonal and a polyclonal origin of plaque smooth muscle cells. We propose that human arteries grow by expanding coherent smooth muscle clones, with little mixing of adjacent clones. Determining whether plaques arise by clonal expansion will require other approaches, such as analysis of somatic mutations; the finding of large X-inactivation patches raises the possibility that plaques arise from a pre-existing (developmental) clone.