The steady-state kinetic mechanism of bovine brain farnesyl:protein transferase (FPTase) has been determined using a series of initial velocity studies, including both dead-end substrate and product inhibitor experiments. Reciprocal plots of the initial velocity data intersected on the 1/[s] axis, indicating that a ternary complex forms (sequential mechanism) and suggesting that the binding of one substrate does not affect the binding of the other. The order of substrate addition was probed by determining the patterns of dead-end substrate and product inhibition. Two nonhydrolyzable analogues of farnesyl diphosphate, (alpha-hydroxyfarnesyl)phosphonic acid (1) and [[(farnesylmethyl)hydroxyphosphinyl]methyl]phosphonic acid (2), were both shown to be competitive inhibitors of farnesyl diphosphate and noncompetitive inhibitors of Ras-CVLS. Four nonsubstrate tetrapeptides, CV[D-L]S, CVLS-NH2, N-acetyl-L-penicillamine-VIM, and CIFM, were all shown to be noncompetitive inhibitors of farnesyl diphosphate and competitive inhibitors of Ras-CVLS. These data are consistent with random order of substrate addition. Product inhibition patterns corroborated the results found with the dead-end substrate inhibitors. We conclude that bovine brain FPTase proceeds through a random order sequential mechanism. Determination of steady-state parameters for several physiological Ras-CaaX variants showed that amino acid changes affected the values of KM, but not those of kcat, suggesting that the catalytic efficiencies (kcat/KM) of Ras-CaaX substrates depend largely upon their relative binding affinity for FPTase.