In crop studies, shading is a common approach for reducing source activity (i.e. assimilates produced by photosynthesis); however, if source reduction is not directly proportional to the degree of light attenuation, estimations could be flawed. We analysed the effects of shading at three levels on light transmittance, air and canopy temperatures, and actual photosynthetic activity at different times of day, and on grain weight in a modern genotype of bread wheat (Triticum aestivum L.). The treatments were imposed 10 days after anthesis and comprised three levels of shading with meshes: 80%, 50% and 10% transmittance during midday. The decrease in midday photosynthetic activity (electron transport rate, ETR; and CO2 assimilation, An) in the shading treatments compared with control plants did not fit a 1 : 1 relationship with light transmittance of the meshes; for example, the reduction in ETR and An was <50% under the mesh with 50% light transmittance. The decrease in light transmittance and photosynthesis caused by the meshes was higher during the morning and afternoon than midday. In the grain-filling period, midday ETR showed a progressive, age-dependent decrease in control plants compared with shading treatments. Initial differences in photosynthetic activity between treatments were attenuated throughout the grain-filling period. The diurnal course of air temperature showed a progressive decrease from control to shade treatments, mainly during midday. This pattern was confirmed by infrared thermometry of canopy temperature, with severe shade showing the lowest temperature. Greenness measurements at 19 days after anthesis showed delayed senescence in shaded plants, possibly due to the decrease in temperature under shade. Grain weight per ear was reduced in shading treatments (mainly with 50% and 10% light transmittance). The decrease in grain weight per ear plotted against the decrease in photosynthesis (evaluated through ETR or An) did not fit a 1 : 1 relationship. This could be explained in part by an observed increase in dry matter retranslocation from the stem. In summary, estimations of source reduction on the basis of light attenuation could be biased, because the photosynthesis decrease is less than proportional, and photosynthesis differences are attenuated throughout grain filling.