There have been a few notable occasions when the Australian wheat segregation system (mainly based on specification of variety and protein content) has failed to produce grain which gives dough properties expected for the wheat grade. The reasons for this are likely to relate to growing and storage conditions; of these, variations in temperature during grain filling appear to be a major factor. Observations of crop statistics, field and glasshouse experiments indicate that as growth temperatures increase up to 30°C, there is a general increase in dough strength (as indicated by Extensograph maximum resistance, Rmax, and as Farinograph development time and stability). However, a decline in dough strength is observed following periods of heat stress (e.g. a few days with maxima of over 35°C). Increasing temperatures during grain filling have also been observed to produce grain with a higher protein content, but this observation is not as consistent nor as marked as the effects on dough strength. We have sought to identify genotypes that do not follow this general trend in response to heat stress, and thus could be used as parents to breed for heat tolerance and greater stability of dough quality. A glasshouse experiment involving 45 genotypes has indicated that there is some variation in the response to heat stress, with a few genotypes being promising sources of tolerance. A second important approach to minimising the effects of heat stress is to develop a model to predict grain-quality changes, thus enabling a marketing authority to be forewarned of significant variation from the quality attributes normally expected for a wheat grade, and assisting breeders to better interpret the results of quality testing of lines grown at various sites.