This retrospective study reviews the results of our experience with the occurrence of CMV DNAemia in islet cell transplanted cynomolgus monkeys subjected to different immunosuppressive protocols, including induction treatment with thymoglobulin (TMG), with a combination of thymoglobulin and fludarabine (FLUD), with cyclophosphamide, or with daclizumab. CMV DNA in the peripheral blood (CMV DNAemia) of 47 monkeys was quantified by real-time PCR on a weekly to biweekly basis. As compared to other immunosuppressive regimens, and in association with greater decreases in WBC, lymphocyte, CD3+CD4+, and CD3+CD8+ lymphocyte counts, frequent CMV DNAemia occurred earlier (within the first month posttransplant), and was of greater severity and duration in recipients of TMG ± FLUD. Treatment of recipients with alternative induction agents that resulted in less dramatic reductions in WBC and lymphocyte counts, however, resulted in occurrence of CMV DNAemia after postoperative day 60. The frequency, average intensity, duration, and area under the curve (AUC) for CMV DNAemia in animals receiving TMG ± FLUD were 75-100%, 4.02 ± 1.75 copies/ng DNA, 23.0 ± 5.3 days, and 367.0 ± 121.1 days × copies/ng DNA, respectively; corresponding values in animals receiving other treatments (0-44%, 0.19 ± 0.10 copies/ng DNA, 0.5 ± 0.3 days, and 75.4 ± 40.2 days × copies/ng DNA, respectively) were significantly different. The value of WBC, T and B cells at the nadir of cell depletion greatly affects the occurrence of CMV DNAemia. No animals developed CMV DNAemia within the next 3 weeks when the lowest value of WBC, lymphocyte, CD3+, CD3+CD4+, CD3+CD8+, or CD20+ cells was above 4500, 1800, 300, 200, 150, or 300 cells/μl, respectively. Oral valganciclovir prophylaxis did not completely prevent the appearance of CMV DNAemia.