Bacterial proteins are synthesized with a formylated methionine (fMet) at the N-terminus. The majority of proteins undergo co-translationally processing of the N terminus by removal of the formyl group and subsequent methionine excision. Deformylation is carried out by peptide deformylase (PDF), which is an essential enzyme in bacteria and important for eukaryotic cell organelles. PDF has been shown to bind to the ribosome, giving rise to the hypothesis that PDF deformylates the N-termini of proteins as they emerge from the peptide exit tunnel of the ribosome. We have developed an in vitro assay that allows us to monitor deformylation of ribosome-nascent chain complexes (RNCs) which are the in vivo PDF substrates. The use of RNCs as substrates has provided insight into the effects of nascent chain sequence, structure, and length on deformylation. PDF substrates have an optimal nascent chain length of about 75 amino acids, and α-helix formation near the N-terminus reduces the efficiency of deformylation. Disruption of the N-terminal α-helix by proline substitution reverses the effect, suggesting that both length and flexibility of the nascent chain determine the frequency by which it reaches the active site of PDF, which is fixed to its binding site near the peptide exit of the ribosome. Interestingly, one of the RNCs tested exposes the signal-anchor sequence of a bacterial inner membrane protein and is an extremely poor substrate for PDF. This raises the possibility that membrane proteins may fail to be deformylated and, thus, retain formylated methionine at the N terminus.