Cytomegalovirus (CMV) historically has been the major infectious cause of d
eath after allogeneic bone marrow transplantation (BMT). During the 1970s a
nd 1980s, clinical trials identified risk factors for CMV disease. During t
he 1980s, new diagnostic tools led to more accurate and earlier detection o
f disease and identified patients with active infection before the onset of
disease. These methods permitted the development of strategies to prevent
infected patients from developing disease, which in the 1990s culminated in
substantial reduction in CMV morbidity and mortality. In seronegative pati
ents, prevention of acquisition of virus from blood product support has pro
ven to be highly effective. In other patients (who are already infected or
whose donor is infected), the use of immunoglobulin and, more recently, sev
eral antiviral prophylactic or "pre-emptive therapies'' have proven to be h
ighly effective. Unfortunately, once antiviral therapy is stopped, infectio
n can recur with subsequent disease. The increasing use of alternate donors
, where the recipient may have slower recovery of protective immune respons
es, and the growing choice of peripheral blood over marrow as stem cell sou
rce, with attendant greater risk for chronic graft-versus-host disease (GvH
D) are changes in transplant practices that have led to more individuals wi
th delayed recovery of protective immune responses. Thus, we have seen a sh
ift from early onset CMV disease (that is now controlled) to late onset dis
ease. New strategies are needed to identify at-risk patients and to prevent
late onset disease.