Winter wheat (Triticum aestivuin cv. Mercia) was grown in a controlled-environment facility under simulated Held conditions at ambient (360μmol mol−1) and elevated (690 μmol mol−1) CO2 concentrations. Some of the plants were shaded to mimic cloudy conditions during three periods of about 20d duration between terminal spikelet and start of grain-fill, giving 16 treatments in all. Elevated CO2, increased grain yield by about 20%, while shading in any period decreased yield, with the greatest effect in the last period, encompassing anthesis. No interactions between these effects were significant for grain yield, but there were complex interactions for mean grain size. Observed effects of shading and elevated CO2 on biomass production were well predicted by a simulation model. Observed effects of treatments on yield could be related to effects on biomass using a simple model which assumes that yield is proportional to biomass production, with coefficients of 0.42 (g grain yield g−1 biomass) for the first two periods and 0.74 for the last period. Wheat models should therefore include developmental changes in sensitivity of yield to biomass production, but biomass changes induced by different CO2 concentrations or light environments can be treated as having equivalent effects on grain yield.