In October 1969 tests made on 1,457 students entering English universities and colleges showed that 57% already possessed antibody to EB virus. The students without antibody in these initial tests were retested seven months later and by then 12% had acquired antibody. In about one-third of them the acquisition of antibody was not associated with any illness. In about 20% respiratory and other illness had occurred, but these symptoms were almost equally frequent in students who had not acquired antibody. Nearly half had developed infectious mononucleosis. In students in whom the acquisition of EB virus antibody was associated with the clinical and haematological manifestations of infectious mononucleosis the Paul-Bunnell test was almost invariably positive. In contrast, when these manifestations were not associated with the acquisition of EB virus antibody the Paul-Bunnell test was always negative. Tests for cytomegalovirus antibody were also made on the students at entry. The proportion of students with this antibody was much less (30%) and only a small proportion (1·4%) of those without antibody had acquired cytomegalovirus antibody seven months later. In the only patient in whom the acquisition of cytomegalovirus antibody alone was associated with the clinical and haematological features of infectious mononucleosis the Paul-Bunnell test was negative.