Why do some genomes evolve fast, some slow? Does abundance determine genetic polymorphism levels? Selection efficiency? What is the proportion of adaptive amino-acid substitutions? Why do some species, but not all, optimize codon usage? What controls base composition variations within and between genomes? Why is the mitochondrial genome hypermutable in animals, but not in plants? Does self-fertilization impede molecular adaptation? These and other basic questions regarding molecular evolution are not answered yet, or only vaguely, despite the huge amount of data available, and the strength of the population genetic theory. One limiting factor is probably the taxonomic bias of current population genomic data sets, especially in animals - most of the evolutionary genomic literature is in mammals or Drosophila. The PopPhyl project aims at characterizing within- and between-species molecular variations in a substantial number of metazoan taxa thanks to next-generation sequencing technology, with the hope of linking genome evolutionary patterns to species biology and ecology.