Cathepsin (Cat) K is a critical bone-resorbing protease and is a relevant target for the treatment of osteoporosis and bone metastasis, while CatS is an attractive target for drugs in autoimmune diseases (e.g. rheumatoid arthritis), emphysema or neuropathic pain. Despite major achievements, current pharmacological inhibitors are still lacking in safety and may have damaging side effects. A promising strategy for developing safer reversible and competitive inhibitors as new lead compounds could be to insert non-cleavable bonds at the scissile P1-P1' position of selective substrates of CatS and CatK. Accordingly, we introduced a 1,4-disubstituted 1,2,3-triazole heterocycle that mimics most of the features of a trans-amide bond, or we incorporated a semicarbazide bond (azaGly residue) by replacing the α-carbon of the glycyl residue at P1 by a nitrogen atom. AzaGly-containing peptidomimetics inhibited powerfully their respective target proteases in the nM range, while triazolopeptides were weaker inhibitors (Ki in the μM range). The selectivity of the azaGly CatS inhibitor (1b) was confirmed by using spleen lysates from wild-type vs CatS-deficient mice. Alternatively, the azaGly bradykinin-derived CatK inhibitor (2b) potently inhibited CatK (Ki = 9 nM) and impaired its kininase activity in vitro. Molecular modeling studies support that the semicarbazide bond of 2b is more favorable than the 1,2,3-triazole linkage of the bradykinin-derived pseudopeptide 2a to preserve an effective affinity towards CatK, its protease target.