Catalytic phenylalanylation of tRNAPhe and the reverse reaction, AMP-dependent deacylation of Phe-tRNAPhe, have been measured by steady-state and pre-steady-state techniques, including rapid sampling and fluorescence stopped-flow methods. (1) Stoichiometry of adenylate synthesis under steady-state phenylalanylation of tRNAPhe indicates half-of-the-sites reactivity. (2) Identity of values of rate constants under pre-steady- and steady-state conditions demonstrates that the rate-limiting steps in catalysis are bond making for phenylalanylation and bond breaking for AMP-dependent deacylation, respectively. (3) Values of catalytic rate constants are the same as those for the conformational change of the Phe site directed enzyme-Phe-tRNAPhe complex [Baltzinger, M., & Holler, E. (1982) Biochemistry (preceding paper in this issue)]. (4) A model is developed that accounts for the observed concert of chemical and geometrical reactions as well as for experimental evidence that nascent Phe-tRNAPhe may not be the same as in solution. In this model, nascent Phe-tRNAPhe is thought to be the tetrahedral intermediate that is formed by nucleophilic attack of the adenylate by the tRNA. It awaits the conformational change in order to break down into Phe-tRNAPhe and AMP. The model can serve as a unifying basis for an interpretation of discrimination against noncognate amino acids and tRNAs and also gives an explanation why severe product inhibition is not observed [Güntner, C., & Holler, E. (1979) Biochemistry 18, 2028-2038].