OBJECTIVE To document injury rates in professional rugby players in the Rugby Super 12 competition and to act as a pilot study for future studies of rugby injuries. METHODS Prospective longitudinal study encompassing the 1997 Super 12 rugby season. METHODS A New Zealand Super 12 rugby squad. METHODS 25 professional rugby players (replacement players were used for unavailable players, so although 30 different players were used during the season, there were only 25 in the squad at any one time). METHODS An "injury" was defined as something that prevented a player from taking part in two training sessions, from playing the next week, or something requiring special medical treatment (suturing or special investigations). An injury was "significant" if it prevented the player from being able to play one week after sustaining it (that is, if it made the player miss the next match). RESULTS The overall injury rate was 120/1000 player hours. The rate of significant injuries was 45/1000 player hours. Those playing the position of "forward" had a higher overall injury rate than other players, but there was no difference in significant injury rate between the forwards and the backs. Injuries that caused players to miss game time occurred almost exclusively during the pre-season program or in the final third of the season. The majority of injuries were musculo-tendinous sprains or strains. The phase of play responsible for the majority of injuries was the tackle. The most frequently injured body part was the head and face. No catastrophic injuries occurred during the study period. CONCLUSIONS Injury rates increase with increasing grade of rugby, injury rates in the Super 12 competition being higher than in first grade rugby. There is very little quality data on rugby injuries, and the few studies available use different methods of data collection and injury definition. There is a pressing need for the collection of accurate ongoing epidemiological data on injuries in rugby.