Distance running continues to experience increased participation in the Western world, although it is associated with high injury rates. Barefoot running has been increasingly proposed as a means to prevent overuse injury due to various biomechanical differences, including reduced joint loading rates and altered kinematics and muscle activity patterns compared to shod running.The aim of this review was to systematically evaluate biomechanical differences between running barefoot and shod, including the quality of available evidence, in order to provide guidance on the phenomenon of barefoot running to the running and sports medicine communities.A comprehensive search of MEDLINE, Web of Knowledge and EMBASE from inception to January 2013 was performed.Trials evaluating injury-free recreational or competitive adults who participate in long-distance running (≥5 km), where a comparison of barefoot and shod running lower-limb kinetics, kinematics and/or electromyography were included. Studies examining sprinting and studies of single-subject design were excluded.Following initial searching, two reviewers identified a shortlist of relevant studies based on title and abstract, with the full text of these studies being tested against the inclusion criteria. References of included studies were examined and citation tracking was performed in Web of Knowledge. Two independent reviewers evaluated the methodological quality of each included study using a modified version of the Downs and Black quality index. Results of the quality assessment were used to identify high- and low-quality studies, data pooling was completed where possible and levels of evidence were determined based on the van Tulder criteria.Eighteen studies were identified, all of low methodological quality. Effect size (ES) calculation was possible for 12 studies. Pooled results indicate moderate evidence that barefoot running is associated with reduced peak ground reaction force (GRF), increased foot and ankle plantarflexion and increased knee flexion at ground contact compared with running in a neutral shoe. Limited evidence indicates barefoot running is associated with reduced impact GRF, reduced peak knee flexion and varus joint moments, and a higher stride frequency compared to a neutral shoe. Very limited to limited evidence also indicates power absorption at the knee is decreased while being increased at the ankle whilst barefoot running. Additionally, the effects of barefoot running on loading rate appear dependent on strike pattern adopted, with a forefoot strike pattern found to reduce loading rate, whilst a rearfoot strike pattern increases loading rate when running barefoot compared to shod.Key methodological weaknesses that must be addressed in future research were identified. Of particular note were absence of investigator blinding, infrequent intervention randomisation, small sample sizes and lack of evaluation following habituation. Two studies could not be retrieved because of publication in a non-English-language journal. Of particular note is that the validity of the body of work is compromised by the lack of evaluation after habituation, or re-training, of previously shod rearfoot-striking runners to barefoot forefoot-striking running styles.There has been a great deal of publicity for barefoot running, and many claims made about its effects and risks. Despite a large amount of biomechanical data available for meta-analysis, clear guidance for clinical practice is limited because of the low methodological quality of the associated studies. Preliminary biomechanical differences identified suggest barefoot running may be associated with positive biomechanical changes in regards to injury prevention, although this may be dependent on strike pattern adopted. Further research employing more robust methodology, which addresses weaknesses highlighted in this review, is needed to confirm current preliminary evidence. Additionally, prospective research would have higher validity were the biomechanical effects of habituating to barefoot running fully examined alongside an evaluation of prevention of repetitive use injury.