Many epidemiological studies have shown positive short-term associations between health and current levels of outdoor air pollution. The aim of this study was to investigate the association between air pollution and the number of visits to accident and emergency (A&E) departments in London for respiratory complaints. A&E visits include the less severe cases of acute respiratory disease and are unrestricted by bed availability. Daily counts of visits to 12 London A&E departments for asthma, other respiratory complaints, and both combined for a number of age groups were constructed from manual registers of visits for the period 1992-1994. A Poisson regression allowing for seasonal patterns, meteorological conditions and influenza epidemics was used to assess the associations between the number of visits and six pollutants: nitrogen dioxide, ozone, sulphur dioxide, carbon monoxide, and particles measured as black smoke (BS) and particles with a median aerodynamic diameter of <10 microm (PM10). After making an allowance for the multiplicity of tests, there remained strong associations between visits for all respiratory complaints and increases in SO2: a 2.8% (95% confidence interval (CI) 0.7-4.9) increase in the number of visits for a 18 microg x (-3) increase (10th-90th percentile range) and a 3.0% (95% CI 0.8-5.2) increase for a 31 microg x m(-3) increase in PM10. There were also significant associations between visits for asthma and SO2, NO2 and PM10. No significant associations between O3 and any of the respiratory complaints investigated were found. Because of the strong correlation between pollutants, it was difficult to identify a single pollutant responsible for the associations found in the analyses. This study suggests that the levels of air pollution currently experienced in London are linked to short-term increases in the number of people visiting accident and emergency departments with respiratory complaints.