In this study, a prototype electronic nose was developed for monitoring the quality of cocoa beans.  The system comprises an array of metal-oxide semiconductor sensors and an artificial neural network pattern recognition unit.  The results obtained from assessment experiments on cocoa beans show good agreement with those obtained from the traditional ‘cut test', recording up to 95% accuracy.  This investigation demonstrates that the electronic nose technique holds promise as a successful technique in evaluating the quality of cocoa beans for industrial processing.