Prior studies have generally suggested that Artificial Neural Networks (ANNs) are superior to conventional statistical models in predicting consumer buying behavior. There are, however, contradictory findings which raise questions about the usefulness of ANNs. This paper discusses the development of a series of artificial neural networks for modeling adoption of online shopping and compares the results to equivalent logistic regression models. Our findings suggest that under simple model assessment criteria, such as Percentage of Correct Classification (PCC) and Area Under Receiver Operating Characteristic (AUROC), ANNs consistently outperform logistic regression. However, findings were, to some extent, inconclusive when more detailed measures (i.e., ROC curve and cumulative lift) were used to compare the competing models. Overall, our results show that ANNs generate predictions that are comparable to, or marginally better than, logistic regression for most of the surveyed products.