Peptide ligands are valuable tools for basic research and drug discovery, and benefit from the availability of powerful discovery platforms based on affinity selection from genetically-encoded libraries. For decades, large libraries of random peptides have also been accessible synthetically (~10^9 peptides), but have generally been examined by screening. Here, I will discuss how de novo sequencing by mass spectrometry has become a viable solution to analyzing peptide mixtures, and enables the application of ‘affinity selection – mass spectrometry’ to synthetic mixtures that approach the diversity of phage libraries. This approach provides a means to examine synthetic libraries of considerably higher diversity than has been feasible historically, creating new opportunities for their use in ligand discovery. Toward this goal, progress will be discussed along three fronts: 1) assessing the scope and reliability of de novo sequencing, such that it can be applied with confidence to unknown mixtures; 2) developing optimal affinity selection procedures, such that binding determinants can be inferred from sequencing output; and 3) developing synthetic libraries that complement existing discovery approaches.