The purpose of this critical review was to propose methodological developments in sport injury prediction and prevention research. Altogether, 24 studies (e.g., quantitative, qualitative, and prevention intervention studies) conducted from 2006 forward were analysed, related to the "stress-injury model." The injury prediction studies were mostly based on prospective designs, using regression analysis, and studied trait anxiety and life stress. The qualitative studies used mainly thematic analysis, and the intervention studies showed some promising effects, but also inconclusive results. We proposed five specific needs for future research: (a) focus on separate research cohorts, (b) variation in preventive intervention designs, including sound protocols conducting experimental studies, (c) focus on behaviours in relation to cognition, (d) application of repeated-measure designs, and (e) use of statistics that could test complex interactions and intraindividual differences. Future research attention should also be oriented towards the psychology of overuse injuries, biopsychosocial perspectives, and health economic evaluations. While progress has been made in research on psychological antecedents of sport injury, prevention, and intervention in the last 10-15 years, several methodological issues still remain to be further developed, as outlined in this article.