We evaluated the utility of plasma polymerase chain reaction (PCR) for surveillance of human herpes virus 6 (HHV-6) infection among pediatric bone marrow transplant (BMT) recipients. We used a prospective, non-interventional design involving a study group and controls. BMT recipients and healthy controls were evaluated. BMT subjects had HHV-6 PCR done biweekly for 12 weeks post transplantation, while a single PCR test was done on controls. For the PCR assay, EDTA blood was collected and DNA extracted from whole blood and cell-free plasma using standard procedures. The PCR was first performed on DNA from whole blood and if a positive result was obtained, the test was repeated on the DNA from the plasma. Thirty BMT recipients (13 autologous and 17 allogeneic) were enrolled, on whom a total of 156 PCR tests were performed, while six tests were done on six healthy controls. The median age of BMT subjects was 6.2 years (range 0.5-17.5 years). The median age of the control subjects was 6.6 years (range 2-10 years). Among asymptomatic BMT patients who had PCR surveillance, the positivity rate was 3.3% (1/30) on whole blood and 0% (0/30) on plasma. None of the six healthy subjects had a positive PCR test on whole blood. During the period of the surveillance study, 14 patients had diagnostic evaluations for HHV-6 disease because of clinical symptoms. Two of these patients were diagnosed with disease associated with HHV-6 (graft failure and encephalitis) and had positive PCR tests on whole blood and plasma and whole blood and cerebrospinal fluid, respectively. We conclude that despite the fact that HHV-6 seropositivity rates are high among children, the frequency of HHV-6 plasma PCR positivity is low in pediatric BMT subjects who are asymptomatic for HHV-6 disease. Given that a positive test on plasma is consistent with active infection, this increases the utility of the PCR test as a diagnostic aid in evaluating syndromes presumed to be due to HHV-6 in pediatric bone marrow transplant recipients.