Background: Central measurement of plasma HIV-RNA (pVL) is common in clinical trials. Despite the increased expense and complexity, central testing is believed to reduce variability. In the setting of two international randomised clinical trials we aim to compare trial outcomes on the basis of central and local laboratory assessed pVL. Methodology: Data were available from ENCORE1 and SECOND-LINE study participants. Central pVL data were generated at a single laboratory from batch retrospective assay of cryo-preserved samples (Abbott m2000rt lower limit of detection (LLD) 40 copies/mL) and 33 local laboratories (LLD 20-50 copies/mL) performing a variety of real time assays under differing infra-structure constraints. The trials’ endpoint, non-inferiority between arms (Δ -10 and -12% respectively) in proportion <200copies/mL at week 48, intention to treat missing=fail were calculated from central+ local (where central was missing) and local measure alone. The distribution of log10 pVL copies/mL at week 0 was summarised and described using a Bland-Altman plot (distribution of difference }1.95 standard deviations (lower (LLA) and upper level (ULA) of agreement) and average of test results). Results: Of the participants in follow up at week 48, 98% had local and central pVL measures (ENCORE1 603/615, 602/615; SECOND-LINE 510/519, 508/519 respectively). The point estimates (% difference [95%confidence interval]) determined by central+local measure were very similar to those by local measure alone in both studies (1.9 [-1.9, 2.1], 2.5 [-1.6, 6.6] ENCORE1; 1.8[-4.7, 8.3], 0.3[-6.4, 7.0] SECOND LINE respectively). Non-inferiority in both trials was established by both determinations. At week 0 mean pVL was greater than 4.0 log10 copies/mL in both trials on both measures. Local measures were marginally lower than central measures (mean difference pVLlog10 -0.10 and -0.17 respectively) (Figure 1). The limits of agreement were narrower for ENCORE1 than SECOND-LINE (}0.66, }1.14 log10¬ copies/mL respectively). Conclusions: The comparison of primary endpoints revealed no difference in interpretation of trial outcomes based on the use of local or central pVL measures. These data indicate that in a clinical trial setting, where the outcome of interest is pVL within routine limits of detection, central laboratory testing is not warranted.