Ubiquitin-dependent proteolysis by proteasomes generally depends upon the conjugation of polyubiquitin chains to lysine epsilon-NH(2) groups within the targeted proteins. However, engineered lysine-less mutants of certain viral and cellular proteins can undergo polyubiquitination at their N-termini. Is N-terminal polyubiquitination a physiologic process, and how many cellular proteins can be targeted for proteasomal degradation through this mechanism? Recent work indicates that the turnover of the endogenous lysine-less human ARF tumor suppressor protein and its mouse Arf counterpart (containing a single, non-conserved lysine residue) is regulated in this manner.