According to cross-sectional studies in sport science literature, decreasing sports activity with increasing age is generally assumed. In this paper, the validity of this assumption is checked by applying more effective methods of analysis, such as longitudinal and cohort sequence analyses. With the help of 20 years' worth of data records from the German Socio-Economic Panel, the development of sports activity over a lifespan is analyzed. According to inadequate cross-sectional analyses, sports activity decreases with increasing age. In contrast, longitudinal analyses show that sports activity increases with increasing age. Furthermore, cohort sequence analyses indicate that not only age but also cohort and period effects influence sports activity. Thus, different methods of analysis reveal opposite results. (Contains 5 tables and 5 notes.)