Microkinetic analysis of heterogeneous catalytic reactions originated in the early 1970s from a program (KINPAK) that was written to predict the performance of an industrial naphta cracker. This program accepted as input a statement of elementary gas phase cracking and radical-induced hydrogen abstraction reactions, together with their Arrhenius parameters. The reactor conditions of temperature, pressure, flow rate, reactor dimensions, composition of feed, heats of formation of reactants and products, heat capacities, heat transfer coefficients, etc. were also included in the input. Mass balance equations were constructed for every component in the reaction, and the resulting matrix of non-linear equations was solved by a highly modified form of Newton-Raphson iteration [1]. Modification of this program to make it applicable to gas-solid heterogeneous catalysis was effected simply by adding a component that did not flow through the reactor - a catalyst. This paper is a review of the application of this methodology in its application to the industrially important reactions, ammonia synthesis and the water gas shift reaction. It was first applied to the synthesis of ammonia, for which the overall potential diagram constructed from single-crystal data was well established [3]. The surprising result obtained from it was that it predicted rates of ammonia synthesis that were five orders of magnitude too low [2]. Incorporating an energy barrier to nitrogen adsorption, for which experimental evidence from catalytic data existed [13], into the potential energy surface resulted in an exact fit between prediction and experiment [2]. (Changing the original values of the pre-exponential terms from those derived from transition state theory to those obtained experimentally in single-crystal experiments improved the quality pf the single-crystal data derived potential to 10% of experiment, thus highlighting the importance of the factors in the calculations [16].) Validation of the requirement for an activated adsorption has been found in the recent experimental determination by temperature programmed adsorption of the energy barrier to nitrogen adsorption on a real catalyst [20]. The program has also been applied to the water gas shift reaction over a Cu/ZnO/Al2O3 catalyst [21-23]. Here, its application demonstrated clearly the non-uniqueness of the mechanism. A previous kinetic analysis using rate expressions had claimed that it ＇proved＇ that the reaction proceeded through a formate intermediate [24]. This application of KINPAK incorporated a redox mechanism, in which water oxidised the copper, producing hydrogen, while CO reduced the copper, producing CO2. An exact fit between prediction and experiment could be obtained. This application also highlighted the interplay between the separate elementary energetics; different combinations of the activation energies of adsorption of hydrogen and of carbon dioxide to produce an adsorbed carbonate species resulted in the same overall activation energy, but totally different adsorbate compositions and totally different overall activities. The methodology imposes a strict rigour on the mechanisms that are proposed and on the limits within which elementary activation energies can be changed, since the overall potential energy surface is fixed by the heats of formation of the reactants and products. (C)1999 Elsevier Science B.V. All rights reserved.