The purpose of this study was to investigate the relationship between structure and injury in basketball players, with the goal of developing equations to predict injury. We examined 45 subjects who were participating in a community center basketball league. Each subject was measured prior to the season for bilateral weight, quadricep girth, calf girth, Q-angle of the knee, dorsiflexion of the ankle, forefoot varus, rearfoot valgus, and true and apparent leg length. All lower extremity injuries causing a game to be missed were recorded during the 16-game season. Average values for bilateral weight, quadricep girth, Q-angle, rearfoot valgus, and leg length measures for the injured players were all larger than the average for the non-injured players by at least one standard deviation. A logistic regression equation using three of the variables correctly predicted the injury status of 91% of the players. The three-variable equation was then used prospectively to predict injury status for members of a small college basketball team. Only one player missed a game due to injury, and the equation identified that player as the most likely to be injured. This study demonstrates a strong relationship between structural measures and lower extremity injury in basketball players.