Background Dancers destined for a career in professional ballet undergo intensive training from a young age and have a reportedly high risk of injury. The extent of this risk is currently unclear as there is a deficiency of high-quality injury surveillance studies in dance.

Objective To determine the incidence, nature and severity of injuries among a sample of elite pre-professional ballet students and to compare these factors across year level and gender.

Design Prospective cohort study.

Setting Three elite pre-professional ballet schools in London, UK.

Participants 266 (107 male) full-time ballet students aged 15–19 years.

Assessment of risk factors Year level and gender.

Main outcome measurements Physiotherapists recorded injury details using a standardised database during the 2009–2010 academic year. Injury definition included ‘any dance injury resulting in medical attention and time lost from dance for >1 day beyond the day of injury’. Incidence rate (IR) was reported per 1000 dance exposures (DE=1 performance, rehearsal or class). Injury severity was reported according to time loss and treatments required. Other details collected included type and anatomical location of injury and date and time of onset.

Results 378 recordable injuries were sustained by 203 (76%) dancers over the study period. The IR was 1.9/1000 DEs. There were no significant differences in IRs between genders or between 1st and 2nd year students. However, 1st and 2nd year students had a significantly lower rate of injury than 3rd years (IR ratio 0.63, 95% CI 0.49 to 0.81 and IR ratio 0.75, 95% CI 0.59 to 0.95 respectively).

Conclusion In comparison with other adolescent athletic populations, participants in this study had a similar injury incidence but a higher risk of injury. Third year students, who perform more than other students, had a higher injury rate than younger dancers. This may suggest that performances carry a higher risk of injury than classes.