In this study termite mound soil was used as part of concrete mixture. This work shows the development of a computational model, based on artificial neural networks for the determination of compressive strength of concrete materials made by replacing the fine aggregate with termite mound soil. The work involves building a multi-layer perception neural network model which uses experimental data obtained from compressive strength test of concrete made from termite mound soil. The compressive strength predictions were compared with predictions from an alternative model based on regression analysis. The results of the study show that for the termite mound soil based concrete the regression model prediction has a correlation coefficient of 0.94402 and a sum of squares error of 0.72867100, while the neural network model prediction has a correlation coefficient of 0.94918 and a sum of squares error of 0.07629460. Generally, the models predicted well, but the neural network model predicted better than the regression model. The result of the study has adequately demonstrated a cheap, simple, very quick and accurate alternative to experimental method of concrete strength determination.