Although Roche COBAS Ampliprep/COBAS TaqMan (CAP/CTM) systems are widely used in sub-Saharan Africa for early infant diagnosis of HIV from dried blood spots (DBS), viral load monitoring with this system is not practical due to nonspecific extraction of both cell-free and cell-associated viral nucleic acids. A simplified DBS extraction technique for cell-free virus elution using phosphate-buffered saline (PBS) may provide an alternative analyte for lower-cost quantitative HIV virus load (VL) testing to monitor antiretroviral therapy (ART). We evaluated the CAP/CTM v2.0 assay in 272 paired plasma and DBS specimens using the cell-free virus elution method and determined the level of agreement, sensitivity, and specificity at thresholds of target not detected (TND), target below the limit of quantification (BLQ) (<20 copies/ml in plasma or <400 copies/ml in DBS), and VL of <1,000 copies/ml, and VL of <5,000 copies/ml. Reported plasma VL ranged from TND, or <20, to 5,781,592 copies/ml, and DBS VL ranged from TND, or <400, to 467,600 copies/ml. At <1000 copies/ml, agreement between DBS and plasma was 96.7% (kappa coefficient, 0.93; P < 0.0001). The mean difference between DBS and plasma VL values was -1.06 log10 copies/ml (95% confidence interval [CI], -1.17, -0.97; P < 0.0001). At a treatment failure threshold of >1,000 copies/ml, the sensitivities, specificities, positive predictive values (PPV), and negative predictive values (NPV) were 92.7%, 100%, 100%, and 94.3%, respectively. PBS elution of DBS offers a sensitive and specific method for monitoring plasma viremia among adults and children on ART at the WHO-recommended threshold of >1,000 copies/ml on the Roche CAP/CTM system.