The incidence of extreme rain is expected to increase with climate change and affect rice productivity in Japan. This study aims to evaluate the impacts of long and heavy rain on Japanese rice total-factor productivity (TFP) by estimating causality functions. We measured rice TFP by using the TörnqvistTheil and Malmquist indexes for dependent variables and predicted, the influences of future temperature and rain on rice TFP by the causality function associated with crop models and a hydrological model based on climate projections from the global-climate model (GCM). The results initially showed no significant differences between Törnqvist-Theil and Malmquist indices in the effects of climate factors, although some differences emerged in the causality of socioeconomic factors. Second, the effects of rain were always negative, and absolute TFP elasticity against rain was lower than temperature via yield and quality, but poorly drained surface water as well as flooding reduced rice TFP by 2.5 to 4.5%. Third, changes in predicted rainfall under future climate change caused annual rice TFP to fluctuate, and an impact of rain on TFP fluctuations exceeded that of temperature via yield and quality. This is due to significant variations in annual rainfall, even though the measured elasticity against rain was low. Based on these findings, the implications for research and policy-making are discussed.