Prenatal maternal stress can adversely affect birth outcomes, likely reflecting effects of maternal stress hormones on fetal development. Maternal stress might also induce behavioural changes, such as dietary change, that might influence fetal development. Few studies have documented relationships between stress and dietary change in pregnancy. We analysed stress and dietary change among 222 pregnant women exposed to the 2011 Queensland Floods. We assessed women's objective hardship, subjective distress and cognitive appraisal of the disaster; changes in their diets and their associations with infants' gestational age, weight, length and head circumference at birth, head circumference to birth length ratio (HC/BL) and ponderal index. Greater objective hardship was correlated with more negative dietary change, skipped meals and skipped multivitamins. There were no direct effects of stress or dietary change on birth outcomes. However, we observed an interactive effect of dietary change and exposure timing on head circumference for gestational age (HC for GA) (p = 0.010) and a similar trend for HC/BL (p = 0.064). HC for GA and HC/BL were larger among children whose mothers experienced negative changes to their diet in early pregnancy compared with later pregnancy, consistent with a ‘head-sparing’ response with early gestation exposure. Further analyses indicated that dietary change mediates the relationship between objective hardship because of the floods and these outcomes. This is the first report of relationships among an independent stressor, dietary change and birth outcomes. It highlights another possible mechanism in the relationship between prenatal maternal stress and child development that could guide future research and interventions.