OBJECTIVES: In elite age-group swimming it is unclear to what degree common assessments of anthropometric, jump performance and front-crawl critical speed (CS) correlate with competition performance. DESIGN: Cross-sectional field study. METHODS: Forty eight elite national-level junior swimmers (22 males, age 16.5±1.2 y, 26 females, age 15.5±1.1 y; mean±SD) completed anthropometry tests, loaded and unloaded countermovement jumps and a series of front-crawl time-trials to determine CS and supra-CS distance capacity (D'). Years from peak height velocity (PHV) predicted from anthropometric data was used as a maturity indicator. Race performances within 3 months of testing were standardised to compare across distances and strokes. Multiple linear regression models were formulated using these data. RESULTS: Loaded jump height, mass, D', PHV and humerus breadth best predicted 100m performance in males (R2Adj=0.88, p<0.001), while loaded jump height, chest depth and sitting height predicted female 100m performances (R2Adj=0.74, p=0.002). Loaded and unloaded jump height, mass, CS and PHV (R2Adj=0.73, p=0.003) and CS and chest depth (R2Adj=0.33, p=0.03) predicted 200m performance in males and females respectively. CONCLUSIONS: Common assessments of power and aerobic capacity in elite junior swimmers explain more variance in competition performance for male than female swimmers, as well as for 100m rather than 200m events. These findings highlight the need to empirically assess testing regimens and suggest new tests in this population may be required.