In the search for better protocols of preemptive therapy of human cytomegalovirus (HCMV) infection in hematopoietic stem cell transplant (HSCT) recipients, we conducted a randomized trial comparing antigenemia with the nucleic acid sequence-based assay (NASBA) for determination of HCMV immediate-early messenger RNA (IEmRNA) as the guiding assay for initiation of pre-emptive antiviral treatment. In the IEmRNA arm, antiviral therapy was started upon IEmRNA positivity confirmed the following day, whereas in the antigenemia arm, therapy was started in the presence of either at least 2 pp65-positive leukocytes/2 x 105 examined or a single positive leukocyte confirmed the following day. In both arms, treatment was stopped upon 2 consecutive negative results. All patients were monitored for 3 months after HSCT. The primary end point of the study was duration of anti-HCMV therapy. On the whole, 80 children (41 in the IEmRNA and 39 in the antigenemia arm), recipients of transplants from either a relative or an unrelated donor, completed the study. No patient developed HCMV disease. In the IEmRNA arm, the incidence of HCMV infection was higher compared to the antigenemia arm (80% vs 51%, respectively, P =.0069), as well as the percentage of treated patients (66% vs 44%, respectively, P =.045). However, the percentage of relapses and treated relapses was comparable in the 2 arms. There was no significant difference in median duration of therapy per patient. Although these data indicate that IEmRNA determination does not offer advantages in terms of treatment duration, it can safely replace antigenemia, while semiautomation is the major advantage of the NASBA procedure.