1Hopital de la Pitie Salpetriere, Paris, France,2Hospital Germans Trias I Pujol, Barcelona, Spain,3Allgemeines Krankenhaus St Georg, Hamburg, Germany,4Universitatsspital, Zurich, Switzerland,5Auguste-Viktoria-Krankenhaus, Berlin, Germany,6Ospedale Civico, Lugano, Switzerland,7Hopital Jean Verdier, Bondy, France,8Glaxo Wellcome Inc, Research Triangle Park, North Carolina, USA, and9HIV and Opportunistic Infections Development Group, Glaxo Wellcome R&D, Greenford Road, Middlesex UB6 0HE, UK Objectives To evaluate antiretroviral efficacy of abacavir (ABC) in antiretroviral-experienced patients, by intensifying current antiretroviral therapy (CART) in patients with stable, detectable plasma HIV-1 RNA. Methods Thirty-two European centres recruited HIV-1 positive patients with ≤ 36 months of CART experience. Patients were randomized to receive either ABC (300 mg twice daily) plus CART (ABC + CART) or ABC placebo plus CART (CART). We assessed efficacy as measured by plasma HIV-1 RNA and CD4+ cell counts and safety at baseline, weeks 2, 4 and every 4 weeks thereafter until week 48. Protocol-defined criteria enabled patients to switch to open-label ABC from week 8 onwards. Results Ninety-two patients with a median plasma of 3.66 log10 HIV-1 RNA copies/mL and a median CD4+ cell count of 408 cells/μL were randomized to ABC + CART and 93 patients with a median plasma of 3.52 log10 HIV-1 RNA copies/mL and a median CD4+ cell count of 411 cells/μL were randomized to CART. From weeks 8–48, 11 (12%) patients in the ABC + CART group and 34 (37%) patients in the CART group switched to open-label ABC. At week 48, significantly more patients on ABC + CART (23/92, 25%) than on CART (5/93, 5%) had plasma ≤ 400 HIV-1 RNA copies/mL (P 5000 copies/mL (4/44, 9%). CD4+ cell counts increased by 102 cells/μL and 57 cells/μL at week 48 for the ABC + CART and CART groups, respectively (intent-to-treat, switch included). ABC addition had minimal impact on the CART safety profile. Conclusions ABC intensification, in CART-experienced patients with low viral loads and limited reverse transcriptase mutations, most of whom had previously been on double-therapy, resulted in a significant and durable plasma HIV-1 reduction and concomitant increase in CD4+ cell count. The presence of M184V at baseline had minimal impact on the efficacy of ABC.