The question of appropriate balance tasks for training in sports is the starting point of this thesis. This question from training practice is treated in the sense of the problem-oriented basic research. Therefor it is integrated in theoretical background, subsequently, investigated in two basic research experiments, and finally, retransferred to practical problems.In this thesis the question of appropriate balance tasks is regarded as a problem of transfer of learning. Thus, it is asked under which conditions practicing a balance task positively influences performance in another task. Initially, the theoretical review reveals two potential conditions. Experiment 1 leads to a third potential condition. First of all, transfer may be determined by the type and the relevance of involved postural synergies. Secondly, transfer may depend on the involved sensory systems. Thirdly, the orientation of attention may affect transfer. Both experiments examine whether these conditions actually influence transfer among different stabilometer tasks. The results confirm, that the type and the relevance of involved postural synergies as well as the type and the relevance of involved sensory systems determine transfer. Whereas, it cannot be confirmed that movement related attention negatively affects transfer. For training practice in sports it is recommended to practice balance tasks in which postural synergies and sensory systems are involved that are relevant in sports competitions, too.