The Viral Activation Transfusion Study (VATS) was a randomized trial that compared leukocyte-reduced transfusions with unfiltered red blood cell transfusions in HIV and cytomegalovirus (CMV) antibody–positive patients with anemia who were undergoing their first blood transfusion. The relations of the baseline qualitative and quantitative polymerase chain reaction (PCR) measures of plasma CMV viremia, HIV RNA, CD4+cell counts, and quality of life in these study subjects were examined. The 511 study subjects had a median CD4+cell count equal to 15 cells/mm3, and 110 (21.5%) had CMV viremia by qualitative assay. In multivariate models, frequency of positive qualitative CMV increased with decreasing CD4+cell counts (p= .04 trend), higher HIV RNA (p< .001), and a history of CMV disease (p< .001). Quantitative CMV PCR were performed on the 110 qualitative assay–positive study subjects. Median CMV viral load was 1780 copies/ml. In multivariate regression models, lower CD4+cell count (p= .03), and a history of CMV disease (p< .001) correlated with the level of CMV load. HIV RNA load and CMV load were not correlated. A lower Karnofsky score was associated with both the presence and quantity of CMV DNA.