BACKGROUND: There has never been a cross-Canada surveillance project to determine the rate ofLegionellaspecies as a cause of community-acquired pneumonia requiring hospitalization and to determine whether there are any regional differences in the rates of Legionnaires' disease in Canada. Anecdotally, Legionnaires' disease is thought to be uncommon in Western Canada.METHODS: From January, 1996 through to October 31, 1997, a prospective study of the etiology of community acquired pneumonia requiring admission to 15 tertiary care hospitals in eight Canadian provinces was conducted. A urine sample from each patient was tested forLegionella pneumophilaserogroup 1 antigen using a commercially available ELISA assay. A culture of sputum or other respiratory specimens for Legionellaceae was carried out at the discretion of the attending physician. Two hundred thirty-four patients had acute and 6-week convalescent serum samples tested for antibodies toL pneumophilaserogroups 1 through 6 using an ELISA method.RESULTS: 28 of the 850 patients (3.2%) had Legionnaires' disease; 18 of 823 (2.1%) were positive forL pneumophilaserogroup 1 by urinary antigen testing. The rate of Legionnaires' disease, based on urinary antigen, at the Halifax site was higher than that at the other sites (seven of 163 patients versus 11 of 660 [P=0.04]). Of the 28 cases of Legionnaires' disease identified using all methods, 11 of 277 patients (3.9%) were enrolled from Western provinces versus 17 of 573 patients (2.9%) from Eastern provinces (P=nonsignificant).CONCLUSIONS: Legionnaires' disease is just as common in Western as in Eastern Canada.L pneumophilaserogroup 1 may be more common in Halifax than at the other sites studied.