Annual crop yield forecasts are necessary for analysis because evaluating climate-change impacts on world food markets requires supply-response functions, including output prices of the prior year. This research was undertaken to develop yield-response functions of the world food model to evaluate climate-change effects by incorporating a crop model into the yield-trend function. Yield-trend functions of rice, wheat, maize, and soybeans were obtained by estimating logistic functions or linear functions with a logarithmic time-trend term and climate variables. Furthermore, temperature and solar-radiation elasticities of yields were calculated using a crop model of the FAO and IIASA. The functions of the maximum rate of gross biomass production and the maximum net rate of CO₂ exchange of leaves in the crop model were modified by introducing cubic spline interpolation and logistic functions. Smoothing these two functions alleviates drastic changes, but reveals small changes in the elasticities of crop yields compared to the kinked functions and these more realistic elasticities can improve the evaluation accuracy of climate-change impacts on crop supply and demand. These variable elasticities of temperature and solar-radiation were inserted into the yield-trend functions, whereupon the global effects of changes in climate variables, including rainfall, were analyzed. The changes in yields obtained using climate variables of two of the four RCP scenarios were compared with the baseline, for which climate variables were fixed. Results of trend analyses show that yields of rice, wheat, maize, and soybeans under RCP8.5 are lower than those under RCP2.6, except for wheat in China. Results of geographical analysis show that climate change can be expected to affect wheat and maize productions in low-latitude countries. Furthermore, results suggest that climate change will depress rice production in sub-Saharan African countries in the 2040s.