Objective: To study the effect of mycophenolate mofetil (MMF) on the decay rate of plasma HIV-1 RNA and the latently infected cellular reservoir in treatment-naive patients starting antiretroviral therapy. Design: Randomized trial. Methods: A group of 19 HIV-1 infected patients (9 with a chronic and 10 with a primary infection) starting a triple antiretroviral drug regimen were randomized to a group with or without MMF. Plasma samples for HIV-1 RNA were taken and HLA-DR-CD4+ T cells were co-cultured for HIV-1 isolation. Slopes of plasma HIV-1 RNA and cellular viral load decay were calculated for the first 14 days and the first 24 weeks of treatment, respectively. Results: The median plasma HIV-1 RNA daily decay rate in chronically infected patients was 0.25 log10 copies/ml [interquartile range (IQR), 0.18-0.30] with MMF and 0.28 log10 copies/ml (IQR, 0.22-0.32) without MMF (P = 0.56); in primary infected patients, it was 0.31 log10 copies/ml (IQR 0.31-0.32) with MMF and 0.32 log10 copies/ml (IQR, 0.26-0.34) without MMF (P = 0.75). The median daily decay rate of latently infected cells was 0.017 and 0.004 infected cells/106 cells in patients with and without MMF, respectively (P = 0.89). The increase in CD4 T cells was comparable between patients with and without MMF. After stopping MMF, there was an increase in the cellular reservoir in six of eight patients. Conclusion: The addition of MMF to a triple class antiretroviral regimen in treatment-naive patients does not significantly increase the plasma HIV-1 RNA decay rate or the decay rate of the latently infected cellular reservoir. © 2004 Lippincott Williams & Wilkins.