When choosing auction mechanisms sellers can decide how much information to reveal to buyers regarding the quality of the goods sold. Using a field experiment in a market for wholesale automobile auctions we are able to measure the effects of information on auction outcomes. We create random assignments of information about quality, and manipulate the availability of information over time. Our results suggest that, as the theoretical literature generally predicts, more information increases expected revenues. Furthermore, by measuring the effects on different quality grades of automobiles it seems like the increase in revenues are due to more competition for any given vehicle. Finally, we quantify the value of gathering information and releasing it to potential buyers in this setup. JEL classifications C93, D44, D82, L15.