The chemical interaction between fly ash and lime in steam cured fly ash-lime compacts was modelled by artificial neural network to predict the free lime remaining in the mixes after the curing period. Process parameters such as the pozzolanicity of the ash samples, their surface areas, unburnt carbon content, curing period and the proportion of lime in the fly ash-lime mixes were taken as the inputs for the model and the free lime remaining in the mix was taken as the output parameter. A generalised feed forward back propagation three layered neural network model was used with tan hyperbolic transfer function at both the input and the output layers with 400 examples. The for training data after 3000 iterations, the mean square error value was found to be the minimum for the prediction mode.
