INTRODUCTION:This study aimed to identify risk factors associated with the development of common lower limb injuries during initial defence training in naval recruits who were enrolled in a randomised trial. METHODS:Three-hundred and six naval recruits were randomly allocated flat insoles (n=153) or foot orthoses (n=153) while undertaking 11 weeks of initial training. Participant characteristics (including anthropometrics, general health, physical activity, fitness and foot characteristics) were collected at the baseline assessment and injuries were documented prospectively. Injury was defined as the combined incidence of participants with medial tibial stress syndrome, patellofemoral pain, Achilles tendinopathy and plantar fasciitis/plantar heel pain throughout the 11 weeks of training. A discriminant function analysis was used to explore the ability of baseline measures to predict injury. RESULTS:Overall, 67 (21.9%) participants developed an injury. Discriminant function analysis revealed that participants who sustained an injury were slightly younger (mean 21.4±SD 4.1 vs 22.5±5.0 years) and were less likely to be allocated to the foot orthosis group (40% vs 53%) compared with those who remained uninjured. The accuracy of these baseline variables to predict injury was moderate (78.1%). CONCLUSIONS:Lower limb injury was not accurately predicted from health questionnaires, fitness results and clinical assessments in naval recruits undertaking initial defence training. However, although not reaching statistical significance, the use of foot orthoses may be protective against common lower limb injuries. TRIAL REGISTRATION NUMBER:ACTRN12615000024549; Post-results.