The mechanical properties of steel are influenced by its microstructure, which is obtained as a result of processing conditions. It is, thus, important to study the effect of these conditions on the microstructure. A three dimensional microstructure model, based on a cellular automata (CA) model, was previously developed. The model is capable of reproducing a number of trends caused by the difference in processing conditions, like temperature and cooling rate, with respect to the microstructure and growth kinetics. However, a single set of nucleation and growth parameters to define the behaviour through a range of conditions have not been found yet. Thus, the goal of this project is to optimise the cellular automata model with a single set of nucleation and growth parameters which makes it capable of predicting the microstructure and growth kinetics at different processing conditions. The CA model uses a simplified calculation of the carbon concentration profile in the austenite. Validation of this simplified calculation, termed as the ‘semi analytical model’, was performed. This was necessary to ascertain that the error obtained in the calculation of the CA model is not a result of this simplification. It was found that the accuracy of the semi analytical model depends on the nature of the transformation. The error increases when the transformation becomes more diffusion controlled in nature. The improvement in the CA model was carried out in terms of its ability to deal with the interaction of the solutes with the austenite-ferrite interface. This, in turn, affects the thermodynamic conditions applicable at the interface. The two extreme equilibrium conditions which are generally defined are the para-equilibrium (PE) and the local equilibrium with negligible partitioning (LENP). The former represents a constrained equilibrium resulting in faster kinetics, while the latter is associated with short range diffusion leading to slower kinetics. In order to account for the slowing down of the kinetics at the end of the transformation, a transition from PE to LENP was proposed. This ‘gradual transition approach’ is based on the interface velocity and accounts for equilibrium states intermediate to that of the PE and LENP conditions. Isothermal austenite-to-ferrite transformation in a dual phase steel, DP600, at temperatures of 625, 650 and 700 deg. C was studied. The general trends with respect to the ferrite transformation rate, final ferrite fraction and ferrite microstructure were predicted by the model using a single set of parameters. However, in order to explain the trend with respect to the ferrite grain size, the parameter describing the amount of edge cells used for nucleation had to be manually adjusted at different temperatures. A greater contribution of edge nucleation was required with decrease in temperature. The fraction of edge cells used for nucleation was 0.3, 0.1 and 0 at temperatures of 625, 650 and 700 deg. C respectively.