The order of substrate addition to tyrosyl-tRNA synthetase from baker's yeast was investigated by bisubstrate kinetics, product inhibition and inhibition by dead-end inhibitors. The kinetic patterns are consistent with a random bi-uni uni-bi ping-pong mechanism. Substrate specificity with regard to ATP analogs shows that the hydroxyl groups of the ribose moiety and the amino group in position 6 of the base are essential for recognition of ATP as substrate. Specificity with regard to amino acids is characterized by discrimination factors D which are calculated from kcat and Km values obtained in aminoacylation of tRNATyr-C-C-A. The lowest values are observed for Cys, Phe, Trp (D = 28,000-40,000), showing that, at the same amino acid concentrations, tyrosine is 28,000-40,000 times more often attached to tRNATyr-C-C-A than the noncognate amino acids. With Gly, Ala and Ser no misacylation could be detected (D greater than 500,000); D values of the other amino acids are in the range of 100,000-500,000. Lower specificity is observed in aminoacylation of the modified substrate tRNATyr-C-C-A(3'NH2) (D1 = 500-55,000). From kinetic constants and AMP-formation stoichiometry observed in aminoacylation of this tRNA species, as well as in acylating tRNATyr-C-C-A hydrolytic proof-reading factors could be calculated for a pretransfer (II 1) and a post-transfer (II 2) proof-reading step. The observed values of II 1 = 12-280 show that pretransfer proof-reading is the main correction step whereas post-transfer proof-reading is marginal for most amino acids (II 2 = 1-2). Initial discrimination factors caused by differences in Gibbs free energies of binding between tyrosine and noncognate amino acids are calculated from discrimination and proof-reading factors. Assuming a two-step binding process, two factors (I1 and I2) are determined which can be related to hydrophobic interaction forces. The tyrosine side chain is bound by hydrophobic forces and hydrogen bonds formed by its hydroxyl group. A hypothetical model of the amino acid binding site is discussed and compared with results of X-ray analysis of the enzyme from Bacillus stearothermophilus.