The importance of crude oil in the world economy has made it imperative that efficient models be designed for predicting future prices. Neural networks can be used as prediction models, thus, in this paper we investigate and compare the use of a support vector machine and a back propagation neural network for the task of predicting oil prices. We also present a novel method of representing the oil price data as input data to the neural networks by defining input economic and seasonal indicators which could affect the oil price. The oil price database is publicly available online and can be obtained from the West Texas Intermediate crude oil price dataset spanning a period of 24 years. Experimental results suggest the neural networks can be efficiently used to predict future oil prices with minimal computational expense.