Recent data have suggested that the activity of human tissue-nonspecific alkaline phosphatase (TNAP) could be targeted therapeutically to ameliorate soft-tissue ossification abnormalities resulting from insufficient production of inorganic pyrophosphate (1). Thus, understanding the mechanism of action and precise binding site for inhibitors of TNAP function is paramount. We compared the modeled three-dimensional structure of TNAP with the 3D structure of human placental alkaline phosphatase (PLAP), and identified the residues that differ between these two isozymes within a 12 Å radius of the active site. We then used site-directed mutagenesis to substitute TNAP residues to their respective homologues in PLAP or in the related germ cell (GCAP) isozyme or to Ala. In addition, we mutagenized most of the corresponding residues in PLAP to their TNAP homologues, and studied the role of a conserved residue Y371 in TNAP using the A371 mutation. All mutants were characterized for their sensitivity towards the uncompetitive inhibitors Lhomoarginine (L-hArg), levamisole, theophylline and L-phenylalanine. We found that the identity of residue108 in TNAP largely determines the specificity of inhibition by LhArg. The conserved Tyr-371 is also necessary for binding of L-hArg. The selectivity of inhibition by L-Phe was determined by the identity of residues 108 and 109 in TNAP (or corresponding residues 107 and 108 in PLAP). In contrast, the binding of levamisole to TNAP is mostly dependent on His-434 and Tyr-371, but not on residues 108 or 109. Substitutions H434E (as in PLAP), H434Q (as in chicken TNAP), H434S (as in the intestinal isozyme), H434G (as in GCAP), or H434A, all lead to a significant decrease in inhibition. The reciprocal E429H mutation in PLAP improved the inhibition by levamisole by 10-fold. The main determinant of sensitivity to theophylline is His-434. The H434E substitution (as in PLAP) causes a 50-fold reduction in inhibition, making TNAP even less inhibited than wt PLAP. The reciprocal E429H mutation in PLAP improves the Ki for theophylline inhibition 30 times, close to the level of wt TNAP. Interestingly, theophylline inhibition in TNAP could be further improved by introducing an E108F mutation. Thus, we have clarified the location of the binding sites for all three TNAP inhibitors and we have also been able to exchange inhibitor specificities between TNAP and PLAP. These data will help us in drug design efforts aimed at discovering novel and better inhibitors of TNAP for clinical use.