This study utilizes the bootstrap rolling window causality approach to investigate the causal linkage between nuclear energy consumption and economic growth for different sample periods. Previous studies were based on assume of either causal or non-causal linkage availability for all sample periods. In order to determine the causality in each period, the bootstrap rolling window causality test is used in this study. The study concluded that a predictive power of nuclear energy consumption on economic growth exists only for Canada. Especially in United Kingdom, the “neutrality hypothesis” is supported strongly.