Population-based studies of hospital usage have been used to identify the ongoing adverse impacts of photochemical air pollutants on respiratory health. In this study we examined the relationship between the number of daily emergency room (ER) visits for respiratory illnesses (25 hospitals) and outdoor air pollution in Montreal, Quebec (June-August, 1989-1990). Air pollutants measured included 1- and 8-h maximum ozone (O3) and estimated particulate matter < 2.5 microns in aerometric diameter (PM2.5). Seasonal and day-of-week trends, autocorrelation, temperature, and relative humidity were controlled for in-time series regressions. Although O3 levels never exceeded the U.S. National Ambient Air Quality Standard (NAAQS) of 120 ppb (maximum day, 106 ppb), statistically significant (P < 0.01) relationships were found between respiratory ER visits for patients over the age of 64 and both 1- and 8-h maximum O3 measured 1 day prior to the ER visit day during the 1989 summer: ER visits were 18.7% higher than average (95% Cl, 6.5-30.9%) for a mean increase of 44 ppb O3 (1-h maximum), and 21.8% higher than average (95% Cl, 9.7-33.8%) for a mean increase of 38 ppb O3 (8-h maximum). There was an association between respiratory ER visits for the elderly and estimated PM2.5 lagged 1 day (0.1 visit/microgram/m3 PM2.5, P < 0.07), but this was confounded by both temperature and O3. The only finding for a reference group of nonrespiratory conditions was an inverse association between ER visits for infants and O3, but this was confounded by weather. These findings confirm the impression that while air quality standards may protect the respiratory health of the general population, this is not the case for susceptible subgroups such as the elderly.