The cytomegalovirus antigenemia assay remains a useful tool for monitoring reactivation among transplant recipients. In this study, we compared protocols using direct lysis of small-volume, whole blood (WB) samples vs peripheral blood leukocyte (PBL)-enriched fractions. Of 363 evaluable samples, 51 (14.0%) were positive by one or both methods. Sensitivity, specificity, and negative and positive predictive values were similar (76%, 99%, 96%, and 95% vs 71%, 100%, 96%, and 100% for the WB and PBL methods, respectively). Stratification of qualitative results by WBC count revealed comparable detection rates by each method, although the total number of positive results from leukopenic samples was significantly lower than from nonleukopenic samples (P= .04). Correlation between quantitative results was high, yet the degree of clinical agreement was suboptimal. We conclude that the small-volume, WB lysis method yields results statistically comparable to that of a PBL fractionation method but with fewer technical steps and less complexity.