Identifying key physiological factors is essential in cycling; however, the unique nature of BMX decreases the validity and transferability of research findings from other cycling disciplines. Therefore, this study highlighted the physical and physiological characteristics of BMX riders that could influence track performance. Fifteen sub-elite BMX riders (male n = 12; age 18.3 ± 3.3 and female n = 3; 17.7 ± 5.7 years) undertook a battery of laboratory tests on three different occasions, including body composition, upper and lower body strength, flexibility, sprint and aerobic capacity measures. On a separate day, participants completed three full lap sprints on an outdoor BMX track. Correlation and multiple linear regression analyses were performed to develop predictive models of performance across the laboratory tests and race time. The final model indicated power to weight ratio, relative back-leg-chest strength and arm span explained ~87% of the variability in finish time (adjusted R2 = 0.87, p < .01). These findings highlighted the importance of a multidimensional approach for developing BMX race performance. Coaches should prioritise these variables in their training programs and selection of future talents. However, further physiological and biomechanical investigation is needed to validate current findings, particularly among elite riders.