In their article, Kemper and Koppes have presented their assessment of the long-term relationship between physical activity (PA) and aerobic fitness (AF). Specifically, their autoregression analysis did not support a significant causal relationship between current PA and subsequent AF as measured by VO 2max. Thus, the authors question how to interpret this unexpected result. Which way does the causal arrow point between PA and AF? This reflection aims to conceptually reframe the causality question, not to resolve its behavioral or physiologic components. This reflection will explore potential sources of unexpected outcomes, analyze dimensions of cognitive difficulty in the context of exercise science, and introduce complex adaptive systems (CAS) as an alternative framework for asking questions and understanding outcomes in exercise research.