Studies of heat-related mortality have been predominantly based on analyses of underlying cause of death as the single indicator of a population's vulnerability to high temperatures. Examination of both underlying and associated causes of death could provide a more comprehensive understanding of the population at risk. This study analyzes the impact of high temperatures on mortality in Sydney, Australia, during the warmer six months (October-March) between 1993 and 2004, using the underlying and associated cause of death due to all-cause, circulatory, and respiratory disease. Some mortality datasets were also divided into two age groups, 0-64 and 65+. A generalized linear model assuming negative binomial distribution was constructed for the daily mortality counts using daily maximum temperature and hourly maximum concentrations of ozone (03) and particulate matter (PM10) as covariates. With the air pollution terms in a model, the change in mortality was estimated to be between 4.5% and 12.1% for a 10 degrees C increase in maximum daily temperature, depending on mortality dataset. When air pollutants were removed from a model, the above mortality percentages changed by -1.1% to 0.9%. When both underlying and associated causes of death were considered, the effect remained the same or became lower. Maximum temperature has been found to have a significant effect on mortality in Sydney, with PM10 and O-3 confounding the association. (C) 2008 Elsevier Inc. All rights reserved.