In a community-based study we investigated the relationships between maternal stress, maternal social supports, family functioning and proneness to acute respiratory illness (AR1) in childhood. 'Prone' and 'not prone' children were identified from the responses to a mail questionnaire sent to the addresses of a randomly selected group of Adelaide children who had been born in 1983. 'Prone' children (n = 255) were defined by a respiratory score (based on frequency and severity of reported symptoms in the preceding 12 months) in the top quintile of the distribution, while 'not prone' children (n = 227) were defined by a score in the bottom 20% of the range. Further information was obtained from a questionnaire administered at a home visit. Maternal stress levels were determined from a combination of major life events, minor life events and psychological distress. Maternal stress was significantly associated with respiratory proneness in a stepwise multiple logistic regression (adjusted odds ratio/high versus low = 3.8; 95% confidence interval 2.0-7.2; p = 0.000), while controlling for the effects of maternal smoking, group child care, early chest illness, number of siblings, breastfeeding, occupation, sex, age, home heating, birthweight and parental history of respiratory illness. Family dysfunction was associated with respiratory proneness in bivariate analyses but not after adjustment for the effects of other psychosocial factors in multivariate analyses. Lack of maternal social support was not associated with having a child who was prone to respiratory illness. These findings raise a number of questions about the nature and direction of the relationship between parental psychological status and child health.