Rubella virus was centrifuged in sucrose density gradients. One of two densities could be ascribed to the virus, depending upon the suspending medium used. The virus was found at a density of 1.16 g/cm(3) after centrifugation for 18 hr in sucrose gradients prepared in distilled water. By contrast, when the sucrose gradients were prepared in tris(hydroxymethyl)aminomethane (Tris)buffer containing ethylenediaminetetraacetic acid (EDTA), the virus was found at a density of 1.18 g/cm(3) after 18 hr of centrifugation. The virus banded at this higher density after only 2 hr of centrifugation when pretreated by overnight incubation in the Tris-EDTA buffer. A kinetic study showed that, in sucrose gradients containing this buffer, the virus gradually migrated as a single peak of infectivity from a density of 1.16 g/cm(3) after 2 hr of centrifugation to the higher 1.18 g/cm(3) density after 18 hr. The density change was shown to be reversible; after the removal of the Tris-EDTA buffer, rebanding of virus harvested at the heavy density resulted in its banding at the lower 1.16 g/cm(3) density. The data indicate that density change could not be explained on the basis of the loss of some component from the virus or on the basis of the failure of the virus to reach equilibrium. However, it is possible that the two densities observed were a reflection of the existence of rubella virus in different hydration states in the presence and absence of Tris buffer containing EDTA.