We evaluated 16 antibody assays for their performance in discriminating recent from established HIV-1 infection. These approaches were based on antigen specificity, quantity, conformation dependence, and avidity/affinity of HIV-specific antibodies. A panel of 41 sera from subjects who had seroconverted in the previous 2-6 months (n = 20) and from subjects with established infection (>1 year, n = 21) were run in each assay. Compared with anti-Gag and anti-Pol responses, quantitative anti-Env antibody levels were initially lower and ultimately higher, resulting in the greatest spread and least overlap between incident and established infection. Quantitative measurement included end-point titer in Western blot, end-point titer or response at a given dilution in solid-phase enzyme immunoassays (EIAs) with recombinant proteins or synthetic peptides, and IgG capture assays that reflect the relative proportion of IgG that is anti-HIV antibody. Focusing on the anti-env response, we measured specific responses to the V3 region of gp120, to the CD4-binding site of gp120, to a peptide corresponding to the immunodominant region of gp41, and to conformation-dependent epitopes of gp120. We also measured antibody affinity for gp41 peptide and the relative avidity for gp120 or gp41 peptide by thermal or urea-elution assays. These assays also discriminated recent from established infection but were not necessarily superior to the quantitative anti-Env assays. Appropriate approaches, based on distinct principles or combination of principles, can be used to develop simple assays for identifying individuals recently infected with HIV-1.