Attendances at the emergency departments of the nine acute care hospitals serving the Vancouver region, with a population of just under a million people, were recorded from July 1, 1984 to October 31, 1986. Of about 25,500 visits a month, 2.7% were for respiratory conditions; and of these, 41.3% were for asthma. Data from 11 air monitoring stations were also tabulated on a daily basis, giving mean maximal hourly values for SO2, NO2, and O3; daily aerosol sulfate measurements from one station were also analyzed, together with daily temperature data and measurements of the coefficient of haze. In 3 consecutive years, a peak in asthma attendances was noted, starting in the last week of September, and continuing for 3 weeks. In these periods, weekly visits for asthma reached 130 patients; during the rest of the year, the weekly visits for asthma varied between 30 and 90. This peak affected children and adults between the ages of 15 and 60, but no increase was seen in those over 60 years. Although pollution levels increased sharply in the fall, a day-by-day analysis showed that the rise in asthma attendances preceded the increase in NO chi and SO2 levels for the region, expressed as the mean of the hourly maxima across all stations. It seems unlikely that a specific pollen is responsible for this, or that house mite replication is the cause. Soya beans are not shipped out of Vancouver. The cause of the peak has not been identified. Variations in emergency visits by day of the week have been recorded; in children and in those aged 15-60, more visits occur on Sundays than on other days, but this does not occur in those over 60. There is no significant variation in environmental data by day of the week. Intercorrelations between environmental variables and emergency visits have been calculated separately for the periods May 1 to October 31, and November 1 to April 31. SO2, NO2, and SO4 are strongly intercorrelated in both periods; ozone is strongly related to temperature, but less strongly correlated to sulfate than is the case in Southern Ontario. In summer, total emergency visits (but not respiratory visits) are strongly correlated with temperature in all age groups (the hotter the day, the more visits occur); but respiratory visits are not related to temperature, ozone, or NO2 levels. However, in the 15-60 age group, asthma and respiratory visits are correlated in summer with SO2 and SO4 levels (P = less than 0.001).(ABSTRACT TRUNCATED AT 400 WORDS)