Understanding and predicting small-grain cereal development is becoming increasingly important in enhancing management practices. Recent efforts to improve phenology submodels in crop simulations have focused on incorporating developmental responses to water stress and interpreting and understanding thermal time. The objectives of the present study were to evaluate data from three experiments to (a) determine the qualitative and quantitative response of wheat (Triticum aestivum L.) and barley (Hordeum vulgare L.) to water stress and (b) ascertain where in space to measure temperature, to provide information required to improve phenological submodels. The first experiment tested the phenological responses of 12 winter wheat cultivars to water stress for two seasons at two sites. The second experiment tested the timing of water stress on spring barley phenological responses for 2 years. In a third experiment, soil near the shoot apex of field-grown spring wheat was heated to 3°C above ambient soil temperature for three planting dates in each of 2 years, to test whether it is better to use soil or air temperature in calculating thermal time. The general response of wheat and barley to water stress was to reach growth stages earlier (i.e. to hasten development). The most significant response was for the grain filling period. Water stress had little effect on jointing and flag leaf complete/booting growth stages. Thermal time to jointing was highly variable across locations. However, thermal time to subsequent growth stages was very consistent both within and across locations. The winter wheat cultivars tested followed this general response across site-years, but inconsistencies were found, suggesting a complicated genotype by environment (G×E) interaction that makes improving phenology submodels for all cultivars difficult. The G×E interaction was most prominent for anthesis (A) and maturity (M) growth stages. Results of heating the soil at the shoot apex depth were completely unexpected: heating the soil did not speed spring wheat phenological development. These results, and others cited, suggest caution in allocating effort and resources to measuring or estimating soil temperature rather than relying on readily available air temperature as a means of universally improving phenology submodels. These results help quantify the response of wheat to water stress and thermal time for improving crop simulation models and management.