The purpose of this study was to compare existing 1 repetition maximum (1RM) bench press prediction equations in National Collegiate Athletic Association (NCAA) Division IA college football players and determine if the error associated with the prediction of 1RM bench press from the National Football League (NFL)-225 test could be reduced through the addition of anthropometric measurements. Anthropometric measures, 1RM bench press, NFL-225 test repetitions to fatigue, and body composition data were collected on 87 Division IA football players (mean+/-SD age 19.9+/-1.3 years; height 182.3+/-7.3 cm; body mass 102.3+/-21.1 kg; % fat 13.9+/-6.7; 1RM bench press 140.5+/-2 6.6 kg; and NFL-225 reps to fatigue 14.1+/-8.0). Hierarchical regression revealed an R=0.87 when predicting 1RM from the NFL-225 test alone, which improved to R=0.90 with the addition of the anthropometric variables: arm circumference and arm length. The following equation was the best performing model to predict 1RM bench press: 1RM (lb)=299.08+2.47 arm circumference (cm)--4.60 arm length (cm)+5.84 reps @ 225; SEE=18.3 lb). This equation predicted 43.7% of subjects' within +/-10 lb of their actual 1RM bench press. Using a crossvalidation group, the equation resulted in estimates of 1RM which were not significantly different than the actual 1RM. Because of the variability that has been shown to be associated with 1RM prediction equations, the use of actual 1RM testing is recommended when this is a critical variable. However, coaches, scouts, and athletes, who choose to estimate 1RM bench press using repetitions to failure from the NFL-225 test, may benefit from the use of the equations developed in this study to estimate 1RM bench press with the inclusion of simple anthropometric measurements.