In this research, an Artificial Neural Network model was developed to predict metal content from drillhole data using the back propagation algorithm. For validation purposes, results between the actual and predicted mineral grades were compared, and regression analysis of the compared results indicate that the predicted mineral grades were in close proximity to the actual grades. The validated model was used to predict mineral grades at unsampled locations in order to determine the feasibility of drilling in those areas. The optimum results obtained from the neural network were fed to geostatistical techniques for developing a geological 3D block model for mine design. The generalized data from the Neural Network show that Artificial Neural Networks can be used to complement exploration activities and is an effective approach for mineral reserve estimation without worrying about spatial variability or other assumptions.