BAFF (B cell activating factor of the TNF family, also known as BlyS and TALL-1), a TNF family cytokine critical for the development and function of B cells, has been reported to bind to three receptors, BCMA (B cell maturation protein), TACI (transmembrane activator and CAML [calcium-modulator and cyclophilin ligand] interactor), and BAFFR (BAFF receptor), but with widely conflicting values for the affinity and selectivity of binding. BCMA and TACI additionally bind APRIL (a proliferation-inducing ligand), the TNF family ligand most homologous to BAFF. Using soluble, monomeric forms of the receptors, we demonstrate that BAFFR binds BAFF with K_D ∼16 nM, while BCMA binds with K_D ∼1.6 μM, indicating a ∼100-fold selectivity for binding to BAFFR over BCMA. APRIL shows the opposite selectivity, binding to BCMA with K_D ∼16 nM while showing no detectable affinity for BAFFR (K_D > 3 μM). The binding of BAFF or APRIL to these receptors is highly sensitive to assay-dependent avidity effects, likely explaining the widely ranging affinity values reported in the literature. Binding of BAFF to BCMA-Fc, a bivalent fusion protein consisting of the extracellular domain of BCMA fused to the hinge and CH1 and CH2 domains of human IgG1, in solution or coated onto an ELISA plate gave apparent binding affinities of ∼0.63 and ∼0.15 nM, respectively, compared to values of K_D(app) ≤ 30 and ∼100 pM for the corresponding BAFFR/IgG1 fusion protein, BAFFR-Fc. The high selectivity of BAFF for BAFFR versus BCMA is thus partly obscured in these multivalent assays. The intrinsically high selectivity inferred from the measurements with monomeric receptor correlates well with in vivo data from knockout mice, providing a possible explanation for the observations that interruption of the BAFFR gene in the A/WySnJ mouse produces a phenotype similar to the BAFF knockout mouse, while the BCMA knockout mouse has no discernible B cell phenotype.