It is generally believed that most T2-weighted (T2) lesions in the central white matter of patients with multiple sclerosis begin with a variable period of T1-weighted (T1) gadolinium (Gd) enhancement and that T1 Gd-enhancing and T2 lesions represent stages of a single pathological process. Lesion probability maps can be used to test this hypothesis by providing a quantitative description of the spatial distribution of these two types of lesions across a patient population. The simplest prediction of this hypothesis would be that the spatial distributions of T1 Gd-enhancing and T2 lesions are identical. We generated T1 Gd-enhancing and T2 lesion probability maps from 19 patients with relapsing-remitting multiple sclerosis. There was a significantly higher probability (P = 0.001) for T2 lesions to be found in the central relative to the peripheral white matter (risk ratio 4.5), although the relative distribution of T1 Gd-enhancing lesions was not significantly different (P = 0.7) between central and peripheral white matter regions (risk ratio 0.6). Longitudinal data on the same population were used to demonstrate a similar distribution asymmetry between new T1 Gd-enhancing and new T2 lesions that developed over the course of 1 year. Alternative hypotheses to explain this observation were tested. We found no spatial difference in the likelihood of development of persistent T2 lesions following T1 Gd enhancement. The relative distribution of T1 Gd-enhancing lesions was shown to be independent of the dose of Gd contrast agent and the frequency of scanning. Our findings suggest that a proportion of the periventricular T2 lesion volume may arise from mechanisms other than those associated with early breakdown of the blood-brain barrier leading to T1 Gd enhancement.