Neural network methodology is used to predict mean particle size resulting from rock blast fragmentation. A blast data base developed in a previous study is used in the current study. A part of this blast data was used to train a single-hidden layer back propagation neural network model for each of the similarity groups obtained in the same previous study. Levenberg-Marquardt algorithm provided the most stable and efficient training out of the four algorithms evaluated. An extensive analysis was performed using a part of the blast data to estimate the optimum number of units for the hidden layer for each similarity group. The remaining blast data are used to validate the trained neural network models. Capability of the developed neural network models is determined by comparing predictions with measured values and predictions based on one of the most applied fragmentation prediction models appearing in the blasting literature. Prediction capability of the neural network models was found to be strong and better than the existing most applied model. Diversity of the blasts data used is one of the most important aspects of the developed models. The developed neural network models are suitable for practical use at mines.