The experiment described here resulted from simulation analyses of climate-change studies that highlighted the relative importance of changes in the mean and variance of climatic conditions in the prediction of crop development and yield. Growth and physiological responses of four old cultivars of winter wheat, to three temperature and two carbon dioxide (CO2) regimes (350 or 700 ppmv) were studied in controlled environment chambers. Experimental results supported the previous simulation analyses. For plants experiencing a 3 ~ increase in day and night temperatures, relative to local long-term mean temperatures (control treatment), anthesis and the end of grain filling were advanced, and grain and dry matter yields were reduced by 27% and 18%, respectively. Increasing the diurnal temperature range, but maintaining the same mean temperature as the control, reduced the maximum leaf area (27%) and grain yield (13%) but did not affect plant development. Differences among the temperature treatments in both phyllochron interval and anthesis date may have resulted from differences between measured air, and unmeasured plant, temperatures, caused by evaporative cooling of the plants. Thermal time (base ---0 ~ calculated from air temperature, from anthesis to the end of grain filling was about 650 ~ d for all cultivars and treatments. Doubling ambient CO2 concentration to 700 ppmv reduced maximum leaf area (21%) but did not infl uence plant development or tiller numbers.