Human immunodeficiency virus type 1 (HIV-1) genetic diversity poses a challenge to reliable viral load monitoring. Discrepancies between different testing platforms have been observed, especially for non-clade-B virus. Therefore we compare, in antiretroviral therapy (ART)-naïve South African subjects predominantly infected with HIV-1 clade-C, three commercially available assays: the COBAS AmpliPrep/COBAS TaqMan HIV-1 Test version 2.0 by Roche (CAP/CTM v2.0), the BioMérieux NucliSens Version 2.0 Easy Q/Easy Mag (NucliSens v2.0) and the Roche COBAS Amplicor HIV-1 Monitor Test Version 1.5 (Amplicor v1.5). Strong linear correlation was observed and Bland-Altman analyses showed overall good agreement between the assays with mean viral load differences of 0.078 log cp/ml (NucliSens v2.0 - Amplicor v1.5), 0.260 log cp/ml (CAP/CTM v2.0 - Amplicor v1.5) and 0.164 log cp/ml (CAP/CTM v2.0 - NucliSens v2.0), indicating lower mean viral load results for the Amplicor v1.5 and higher mean readings for the CAP/CTM v2.0. Consistent with observations following previous comparisons of CAP/CTM v2.0 versus Amplicor v1.5, the CAP/CTM v2.0 assay detected low-level viremia (median 65 cp/ml) in more than one-third of those in whom viremia had been undetectable (<20 cp/ml) in assays using the NucliSens platform. These levels of viremia are of uncertain clinical significance but may be of importance in early detection of ART resistance in those on treatment. Overall the three assays showed good comparability of results but with consistent, albeit relatively small, discrepancies for HIV-1 clade-C samples, especially in the low-viremic range that should be taken into account when interpreting viral load data.