Equestrian sports are considered high in risk, and successful athletes must achieve resilience to recover from adversities inherent to their riding career. The objective of this study is to identify which factors present in show jumping riders are related to resilient abilities, and how they may be improved. 101 British competitive show jumping riders were asked to complete a questionnaire about their riding experience and to respond to two tests: (1) test of performance strategies (TOPS 3), which analyses the use of self-talk, emotional control, automaticity, goal setting, imagery, activation, relaxation, negative thinking and attention control; and (2) Connor-Davidson resilience scale (CD-RISC-10), which evaluates resilient abilities. Where available, competition results were also analysed. Resilience scores had positive correlations with ‘emotional control’ (r=0.518; P<0.001), and ‘emotional control’ had a positive predictive effect on resilience (χ2=8.508; P<0.005). There were indications of positive correlations between resilience and ‘activation’ (r=0.497; P<0.001), ‘automaticity’ (r=0.437; P<0.001), and of a negative correlation between resilience and ‘negative thinking’ (r=-0.416; P<0.001). Overall CD-RISC-10 mean value was 29.96. ‘Activation’ and ‘automaticity’ were more applicable for riders of higher perceived levels of skill. Responses suggest that ‘activation’ was employed more frequently by male than female riders (Z=-2.118; P<0.05), and that ‘negative thinking’ was more frequent in female than male riders (Z=-1.969; P<0.05). More significant correlations between performance strategies were found in professional (19) than in novice riders (5). Results indicate that the use of performance strategies may be beneficial in the development of resilience. In higher level riders, the development of one set of strategies tends to lead to the development of other strategies. The findings support that the ability to recover successfully from adversities may not be merely inherent, but also under individual control.