Objectives: Using a prospective cohort study design, to establish the incidence and nature of time-loss injuries in English community rugby and to assess the differences between different playing levels. Setting: English community rugby clubs. Participants: Injury information for 4635 matches was collected during seasons 2009/2010 (46 clubs), 2010/2011(67 clubs) and 2011/2012 (76 clubs). Clubs were subdivided into groups A (semiprofessional), B (amateur) and C (recreational) for analysis. Primary and secondary outcome measures: Any injury resulting in 8 days or greater absence frommatch play was reported by injury management staff at the clubs. The primary outcome measure was injury incidence (per 1000 player match-hours) and the secondary outcome measure was severity (ie, days absence). Results: Overall match injury incidence was 16.9 injuries per 1000 player match-hours. Incidence was higher for group A (21.7; 95% CI 19.8 to 23.6) compared with group B (16.6; 95% CI 15.2 to 17.9) and C (14.2; 95% CI 13.0 to 15.5, both p<0.001). The mean time-loss was 7.6 weeks absence, with knee and shoulder injuries the most severe with mean absences of 11.6 and 9.3 weeks, respectively. Half of all injuries occurred to the lower limb, with knee and ankle joint/ ligament injuries the most common diagnoses. Shoulder joint/ligament injuries were the most common and severe upper limb injuries. Contact events accounted for 80% of all injuries and tackles accounted for 50%. Running was the most common non-contact injury event, of which 56% were hamstring injuries. Conclusions: More time-loss injuries occur at higher levels of community rugby. Injury prevention strategies should focus on good technique in the tackle and conditioning exercises for the knee, ankle, hamstrings and shoulder.