This article tests the efficient-markets hypothesis by looking at profits in National Football League (NFL) betting markets. The author tests whether successful betting strategies exist when points scored and allowed earlier in a season can outperform the betting line in predicting the margin of victory in NFL games and finds that profitable strategies exist. In addition, the author finds that over the course of a season, bettors do imperfectly incorporate information about team strength and that NFL victory margins are a highly variable process.