Dynamic balance is vitally important for most sports and activities of daily living, so the assessment of dynamic stability has become an important issue. In consequence, a large number of balance tests have been developed. However, it is not yet known whether these tests (i) measure the same construct and (ii) can differentiate between athletes with different balance expertise. We therefore studied three common dynamic balance tests: one-leg jump landings, Posturomed perturbations and simulated forward falls. Participants were 24 healthy young females in regular training in either gymnastics (n = 12) or swimming (n = 12). In each of the tests, the participants were instructed to recover balance as quickly as possible. Dynamic stability was computed by time to stabilization and margin of stability, deduced from force plates and motion capture respectively. Pearson’s correlations between the dynamic balance tests found no significant associations between the respective dynamic stability measures. Furthermore, independent t-tests indicated that only jump landings could properly distinguish between both groups of athletes. In essence, the different dynamic balance tests applied did not measure the same construct but rather task-specific skills, each of which depends on multifactorial internal and external constraints. ... mehrOur study therefore contradicts the traditional view of considering balance as a general ability, and reinforces that dynamic balance measures are not interchangeable. This highlights the importance of selecting appropriate balance tests.