RESUMO
Recent determinations of the abundance of the light-element Li in very metal-poor stars show that its intrinsic dispersion is essentially zero and that the random error in the estimated mean Li abundance is negligible. However, a decreasing trend in the Li abundance toward lower metallicity indicates that the primordial abundance of Li can be inferred only after allowing for nucleosynthesis processes that must have been in operation in the early history of the Galaxy. We show that the observed Li versus Fe trend provides a strong discriminant between alternative models for Galactic chemical evolution of the light elements at early epochs. We critically assess current systematic uncertainties and determine the primordial Li abundance within new, much tighter limits: &parl0;Li&solm0;H&parr0;p=1.23+0.68-0.32x10-10. We show that the Li constraint on OmegaB is now limited as much by uncertainties in the nuclear cross sections used in big bang nucleosynthesis (BBN) calculations as by the observed abundance itself. A clearer understanding of systematics allows us to sharpen the comparison with 4He and deuterium and the resulting test of BBN.