We compared emergency room visits for the treatment of asthma in two large downtown teaching hospitals: one in Auckland, New Zealand and one in Toronto, Canada. We wished to determine whether the differences in asthma mortality between New Zealand and Canada were reflected in different patterns of emergency room use or physician management. Emergency room use during the past decade was enumerated in both hospitals, and charts containing the sole diagnosis of asthma were reviewed in detail for a defined study period in 1986. In both Toronto and Auckland, the number of emergency visits for asthma had increased significantly in the past decade (P less than .015 but the rate of rise was significantly higher in Auckland (P less than .05). In Auckland, 27% of asthmatics were admitted whereas in Toronto significantly fewer (16%) were admitted (P less than .0005). Objective measures of pulmonary function were documented more frequently by emergency room physicians in New Zealand than in Canada (90% versus 48%; P less than .0005). Pulmonary function measurement was primarily by peak flow meter in Auckland and most commonly by spirometer in Toronto so that pulmonary function measurements could not be compared directly between centers. In both centers, however, admitted patients had significantly lower pulmonary function indices than discharged patients. In New Zealand, mean peak flow was 38% of the predicted value among all asthmatics assessed; in Toronto, mean FEV1 was 47% of predicted. In Toronto, pulmonary function measurements were most likely to be missing among presumably healthier discharged patients. Pulse rate, respiratory rate, and pulsus paradoxus were documented more consistently in Auckland than in Toronto.(ABSTRACT TRUNCATED AT 250 WORDS)