The purpose of this study was to compare the peak periods (1- to 6-minute epochs) for three different training methods (game-based training - GBT; small-sided games - SSG; and conditioning training - CT) in elite male rugby union (RU) players. The peak movement (m·min −1 ) and impact (impact·min −1 ) characteristics of 42 players during in-season training were assessed. When comparing between training methods, SSG drills produced the greatest peak movement characteristics for all time epochs (1-minute average peak periods - SSG 195 m·min −1 , GBT 160 m·min −1 , and CT 144 m·min −1 ). The peak impact characteristics performed during training were 1-2 impact·min −1 for a 1-minute period and then decreased as the time period increased for all training methods. The greatest distribution of training time occurred at 30-39% (SSG and CT) and 40-49% (GBT) of peak movement intensity, with less than 5% of training performed at or above 80% peak intensity across all drill types. Findings from the current study show that the peak movement periods (m·min −1 ) in RU training from all three training methods match or exceed those which are previously reported in peak gameplay, yet their ability to replicate peak impact characteristics is questionable.