This study evaluates HIV infant diagnosis on DBS using Biocentric HIV1-DNA and HIV1-RNA assays, in field conditions in Côte d'Ivoire. Paediatric screening was offered to children ≤ 3 years in clinical sites in Côte d'Ivoire in 2008. For each HIV-infected child, two non-infected children were included and blood samples were collected. HIV-DNA results obtained on EDTA blood samples with Biocentric assay were the reference for HIV infant diagnosis. Plasma and DBS viral loads were measured using HIV-RNA Biocentric assay. DBS samples were also tested for HIV-DNA detection using both Biocentric and Amplicor Roche assays. Sensitivity, specificity and concordance between tests were calculated. Overall samples from 138 HIV-exposed children, 46 infected, 92 non-infected were included. All tests were 100% sensitive and specific including 100% concordance with the two HIV-DNA assays. The median level of HIV-DNA on EDTA samples was 3.15 log10 copies/10(6) PBMCs; the median level of HIV RNA in plasma and DBS were respectively 5.82 and 5.17 log10 copies/ml (Pearson's correlation R(2)=0.92, p<0.0001). The threshold for detectable HIV-RNA on DBS was 3.3 log10. Although there are differences between viral load measured on DBS and plasma, the two Biocentric assays present very good performances for HIV infant diagnosis on DBS while cheap and feasible.