The earliest phases of the chemical evolution of our Galaxy are analysed in the light of the recent VLT results (concerning abundance patterns in the most metal-poor stars of the Galactic halo) and of stellar nucleosynthesis calculations. It is argued that: 1) the unexpected abundance patterns observed in Pop. II stars are not the imprints of an early generation of supermassive Pop. III stars; 2) among the various suggestions made to exlain the observed abundance patterns, nucleosynthesis in asymmetric supernova explosions appears most promising. In the latter case, an indirect correlation between asymmetry and metallicity is suggested by the data. Finally, the VLT data confirm two old ``puzzles'': the existence of primary N early in Galaxy's evolution (which constrains the mixing of protons with He-burning products in massive stars) and the absence of dispersion in abundance ratios, at least up to the Fe peak, in the early Galaxy (which bears on the timescales of homogeneisation of the interstellar medium, but also on yield variations among massive stars).

10 pages, 4 figures, Review talk in Nuclei in the Cosmos VIII (Eds. L. Buchmann et al.) to appear in NuclPhys A