31P NMR signals from substrates and products of alkaline phosphatase have been adapted to measure the rates and product ratios for the hydrolysis and phosphotransferase reactions from pH 6 to 10. Below pH 8, glycerol is a poorer acceptor than H2O (glycerol phosphates:Pi = 0.5). Tris is a more effective acceptor below pH 8, showing a maximum acceptor efficiency at pH 8 (Tris phosphate:Pi = 2). Phosphotransferase efficiencies are in the order expected for the pKaS of the alcohol groups, Tris less than glycerol Cl, C3 less than glycerol C2. Tris and glycerol induce chemical shifts in 113Cd(II) present at the A site but not the B or C sites of the metal triad present at each active center of Cd(II)6 alkaline phosphatase, suggesting that the alcoxides of the acceptors coordinate the A site metal and become the nucleophiles attacking the phosphoseryl residue (E-P) in the second step of the mechanism. The interaction is through the oxygen of Tris. The transferase activity of the amino alcohol shows a bell-shaped pH dependency. Aliphatic alcohol acceptors show small increases in acceptor activity between pH 6 and 8, with 5-fold increases from pH 8 to 10 (at pH 10, glycerol phosphates:Pi = 2.5). 31P NMR inversion transfer has been used to measure the koff for Pi dissociation from the noncovalent enzyme complex (E . P). For the Zn(II)4 alkaline phosphatase koff is essentially pH independent at approximately 35 s-1. For Cd(II) or Mg(II) at the B site in place of Zn(II), koff less than or equal to 1 s-1 X Cl-ion, which appears to coordinate the A site metal ion, enhances koff, suggesting that both Cl- and HPO2-4 can coordinate the A site metal ion in a 5-coordinate intermediate. pH control of the alkaline phosphatase mechanism appears to reside in the stability of E-P and not the dissociation of E . P, compatible with the hypothesis that the activity-linked pKa is that of a H2O molecule coordinated to the A site metal, which in the hydroxide form becomes the nucleophile attacking the phosphoseryl group (E-P).