BACKGROUND.: We assessed the impact of a preemptive strategy after discontinuation of antiviral prophylaxis in the prevention of late-onset cytomegalovirus (CMV) disease in a cohort of kidney transplant recipients. METHODS.: Patients undergoing kidney transplantation at the University Hospital of Lausanne (CHUV) between November 2003 and November 2007 were included if they were donor or recipient (D/R) seropositive for CMV. All patients received 3 months of prophylaxis with valganciclovir, followed by monitoring of CMV DNAemia by polymerase chain reaction (PCR) every 15 days during 3 additional months. Valganciclovir was restarted if CMV PCR was more than or equal to 10,000 copies/mL. The primary endpoint of the study was the incidence of late-onset CMV disease. RESULTS.: Eighty-six kidney transplant recipients were included; 30 patients were D+/R- and 56 patients were R+ for CMV. At 6 months posttransplant, CMV DNAemia had occurred in 31 of 86 (36%) patients: 13 of 30 (43%) in the D+/R- group and 18 of 56 (32%) in the R+ group (P=0.35). In the D+/R- group, among the 13 patients with CMV DNAemia, 7 (54%) patients developed late-onset CMV disease, simultaneously to the first positive viral load (n=5) or after detection of low-grade viremia (n=2). Only two patients received a preemptive treatment. In the R+ group, all positive PCR results were below the established cutoff. Thus, these 18 patients were not treated, and none of them developed late-onset CMV disease (R+ vs. D+/R-: P<0.001). CONCLUSIONS.: Within the limitations of a noncontrolled study, our data indicate that a preemptive strategy after 3 months of valganciclovir prophylaxis for CMV is not useful in R+ kidney transplant recipients. In D+/R- patients, this approach should be further evaluated.