Empirical research on the value of information technology investments in the information systems literature has primarily focused on the use of IT by businesses and multinational firms. The impact of IT on the global agricultural supply chain has largely been ignored in the IS literature. Auctions to buy and sell large volumes of agricultural commodities are widely prevalent in diverse regions of the world and are an important part of the agricultural supply chain. In an effort to increase efficiency, commodity auctions have been experimenting with online formats in recent years. Such online auctions have generated significant interest in the trade press because of their potential to generate higher commodity prices for producers, reduce unfair trading practices by middlemen, and bridge the digital divide. We analyze transaction data from a recently set up online auction in India that trades in various grades of coffee. We model the impact of lower transaction costs, daily operations, less collusive behavior among buyers, and learning curve effects on the selling price of coffee in the online auction. We estimate the parameters of the model by comparing the prices in the electronic auction with those of the same grade of coffee at physical auctions held weekly. We find that electronic auction prices are 4 percent higher and the difference is statistically significant. Further, we find that the price differential is higher for coffee grades that have higher price volatility and that are traded less frequently in the physical exchange. We also find that the price differential increases over time as buyers become more familiar with the benefits of the electronic trading format.