Objectives: The purpose of this study was to document and analyse injuries sustained in premier grade rugby union over a competitive season and investigate the seasonal trend of injury incidence. Design: A prospective epidemiological cohort study of injury. Setting: Field-based collection of match-play injury data. Participants: Two-hundred and seventy-one players from eight premier grade rugby union teams. Main outcome measures: Injury incidence as a function of exposure and match round including descriptive statistical analysis of injury characteristics. Results: Injury incidence during the season was 52 injuries per 1000 player-match hours (95% CI: 42–65). Poisson regression demonstrated a significant decrease in injury rate by 2% for each successive round throughout the season (p < 0.04). Most injuries were sustained during the tackle resulting in soft tissue injuries to the lower limb. Conclusions: The results of this study demonstrate an early season bias of injuries. The majority of injuries were classified as ‘slight’ with players returning to training or play within two days. The tackle was the phase of play which produced the most injuries consistent with previous research. Compared to analogous data collected 10 years previously, injury incidence of a similar cohort was considerably reduced.