Résumé : Objectives: To estimate the prevalence and the evolution over time (1995-1998) of genotypic resistance to antiviral drugs in antiretroviral drug-naive HIV-1-infected patients in Belgium. Design: Belgian Aids Reference Laboratories provided retrospective samples and clinical data from antiretroviral drug-naive HIV-1-infected patients who visited the hospital for the first time in 1995 (n=45), 1997 (n=75) and 1998 (n=111). Genotypic resistance to the three available classes of drugs was monitored using the Line Probe Assay (Innogenetics, Gent, Belgium). Additionally, ARMS-151 was performed for scoring multinucleoside resistance. Results: The prevalence of genotypic resistance at baseline to nucleoside analogue reverse transcriptase inhibitors (NRTIs) and non-nucleoside reverse transcriptase inhibitors (NNRTIs) were each between 10% and 20% for 1995, 1997 and 1998 without an increasing trend over time. For NRTIs, resistance mutations were mainly related to zidovudine in 1995, whereas in 1997 and 1998 baseline resistance was scored for zidovudine, lamivudine or for both drugs simultaneously. No patients displayed the multi-nucleoside resistance Q151M mutation. Baseline resistance mutations to protease inhibitors (Pls) did not rise significantly: 4.4% in 1995, 8% in 1997 and 9.9% in 1998. When scoring any resistance-related mutation, 26.6% displayed genotypic baseline resistance in 1995, 26.6% in 1997 and 31.5% in 1998. Discussion: The prevalence of genotypic baseline resistance to any drug, as scored with LiPA, in naive HIV-1 patients in Belgium is 29%, with baseline resistance mutations to one or several drugs from all available classes of antiviral drugs. The ability of LiPA to pick up minor variants could be an explanation for the higher overall prevalence we observe, when compared to recent estimates in other countries of 16.3% and 22%, which were based on sequencing methods. According to the European guidelines for resistance testing, resistance testing in Belgium before starting antiviral therapy should be considered.