The concept of reducing cell-associated blood-borne viruses (BBVs) by filtration of the vector leukocytes from blood collected for transfusion has led to the development of high efficiency filters. Improved filtration technology demands newer methodology to accurately estimate the residual cells. We have developed an experimental model based on the hemocytometer counts and the polymerase chain reaction (PCR), performed on the lymphocytes derived from the units of red cell mass inoculated with marker cells (H9) persistently carrying cell- associated human immunodeficiency virus DNA (CA-HIV). We measured the efficiency of 6 units of a prototype filter using our model and found an estimated mean of less than 4 residual cells per milliliter in the filtered blood. This represents a mean 5.84 log10 reduction of normal PBMC and CA-HIV in pre- and post-filtration aliquots and exemplifies the application of our model for evaluating a new generation of blood filters. Our model illustrates that a biological tracer (ie, DNA) is a better measure of the efficacy of a leukocyte filter than the hemocytometric enumeration of pre- and post-filtration PBMC concentrates.