Papain-like lysosomal cysteine proteases are processive and digestive enzymes expressed in organisms from bacteria to humans. Their ubiquity alone makes them potential drug targets, with the assumption that appropriate specificities may be achieved. These enzymes have rather short active-site clefts, comprising three well defined substrate-binding subsites (S2, S1 and S1') and additionally have comparatively broad binding areas (S4, S3, S2', S3'). This geometry distinguishes them from other protease classes, such as serine and aspartic proteases, with six and eight substrate-binding sites, respectively. Exopeptidases (cathepsins B, C, H and X), in contrast to endopeptidases (such as cathepsins L, S, V and F), possess structural features that facilitate binding of N- and C-terminal groups of substrates in the active-site cleft. Other than a clear preference for free chain termini in the case of exopeptidases, the substrate-binding sites exhibit no strict specificities. Instead, their subsite preferences arise more from specific exclusions of substrate type. This presents a challenge for the design of inhibitors to target a specific cathepsin: only the cumulative effect of an assembly of inhibitor fragments can produce the desired result. The small number of papain-like lysosomal cysteine proteases (11 human enzymes are known) and the small number of substrate-binding sites calls for a innovative and empirical approach.