Not many years ago most discussion of mutation induction by physical and chemical agents concentrated on the initial lesions induced in the DNA with the implicit assumption that once the lesions were made they were converted almost automatically to mutations by relatively simple processes associated with DNA replication. The discovery of a variety of enzymatic processes that can repair these lesions, the great increase in our understanding of the molecular steps involved in repair, replication, and recombination, and the increasing availability of cells with genetic defects in these proΒ cesses have led to the realization that mutation induction is a far more complex process than we originally thought. Repair systems can remove lesions before they can be converted to mutation, they can also convert initial lesions to secondary ones that are themΒ selves mutagenic, and they can remove potentially lethal lesions at the expense of making mutations. The error-avoiding systems assoΒ ciated with replication are themselves complex and may be caused to make mistakes in various ways. These different pathways for mutation production and mutation avoidance are still being worked out in prokaryotes and are less well understood in eukaryotes. This symposium shows, however, that very encouraging progress has been made in the last several years, and the progress is now accelerating.