With the exception of spinal cord injuries, the frequency and consequences of rugby injuries are not clearly understood. We therefore conducted a prospective cohort study involving all the senior rugby clubs in the Scottish Borders. 1169 (96%) of 1216 eligible players provided personal details and recorded all 15-a-side matches they played in the 1993-94 rugby season. Physiotherapists visited the clubs weekly to liaise with volunteer linkmen who were appointed to notify the circumstances of all new or recurrent injuries occurring in matches or rugby-related training. 361 players experienced 584 injuries in 512 injury episodes, 84% of which arose in matches. The period prevalence rate of match injuries was 13.95 (95% CI 12.64-15.26) per 1000 playing hours, the equivalent of an injury episode every 1.8 rugby matches. An injury episode took the player away from the game for an average of 39 days. 28% of injury episodes resulted in absence from employment or school/college work, for an average of 18 days. Rugby injuries are an important source of morbidity in young men. They need to be better understood if their frequency and consequences are to be reduced.