Gas-solid reactions in chemical and metallurgical industries often involve solid pellets and a gaseous reactant. The progress of chemical reaction is measured by the movement of zones within the pellet and has been explained in terms of diffusion and chemical reaction processes. Earlier models identified a single reaction zone, in addition to product layer and unreacted core. In the present article, two reaction zones are envisaged as a more plausible explanation of the movement of the zones as the reaction proceeds. Earlier models for reversible reactions have assumed that conditions at the interface between the reaction zone and the unreacted core correspond to equilibrium at the prevailing temperature. The gaseous concentrations were assumed to permeate the core at the interfacial values so that no reaction occured in the core. More realistically, the present article envisages an additional zone within which the gaseous concentrations fall from the equilibrium values to zero. It is assumed that in the reaction zone proper, referred to as zone I, having thickness z I , the concentration profile is sigmoidal. This agrees with the earlier work of Khan and Bowen [1] and Prasannan and Doraswamy.[2] In zone I and the concentration of the reactant gas varies between [A i ] and [A*]. In the zone II, having thickness z 2 , concentration varies linearly between [A*] and zero. This model has been applied successfully to the data of the reduction of hematite [3] at different temperatures. The contribution of different forms of resistance, diffusion in product layer, chemical reaction and diffusion in the reactant core, is assessed as function of time (start to the end of reaction). The thickness of the zones remain almost constant as the reaction progresses. In particular, the influences of the product and core diffusion coefficients and chemical equilibrium constant on the extant reaction are evaluated. The dependence of concentration profile and zone thickness on equilibrium constant, K, velocity constant, k, diffusional coefficients D C and D P has been investigated thoroughly. The thickness of both zones has been evaluated for leading variables.