Molecular biology has contributed to the construction of the neoDarwinian synthesis, as it is articulated in the 1980's, in two ways. First, it has provided a nearly complete understanding of the processes of DNA replication and expression, and mutation. We know in detail how genotypes are passed on and how random mutation can provide the variation which must be present in populations if natural selection is to “cause” evolution. (We do not yet understand in detail how genotype determines phenotype in any complex multicellular organism, but there is probably no barrier other than complexity in our way.) Second, molecular biology has provided us with a data base from which to draw conclusions about phylogenetic relationships between species, or between genes, and a powerful way of approaching that data base. Zuckerkandl and Pauling's (1965) contention – that comparisons of amino acid or nucleotide sequences of homologous proteins or nucleic acids from contemporary species should allow measurements of phylogenetic distance – has been amply confirmed. Earlier chapters in this volume provide some details about methods and their application to major groups of organisms. Here I would like to discuss what molecular sequence data, and phylogenetic conclusions drawn from them, might have to say about the very earliest stages of Darwinian evolution, before there were cells.