BACKGROUND--The prevalence and morbidity of asthma vary greatly among different ethnic communities and geographical locations, but the roles of environmental and genetic factors are not fully understood. The differences in prevalence of adult asthma among Chinese, Malay, and Indian ethnic groups in Singapore were examined, and the extent to which these could be explained by personal and environmental factors were investigated. METHODS--A stratified disproportionate random sample (n = 2868) of Chinese (n = 1018), Malays (n = 967), and Indians (n = 883) of both sexes was drawn from households in five public housing estates, and an interviewer administered questionnaire was used to determine cumulative and current prevalence of "physician diagnosed asthma" (symptoms with a physician diagnosis of asthma). RESULTS--Lifetime cumulative prevalence (standardised to the general population) of "physician diagnosed asthma" was 4.7% in men and 4.3% in women; 12 month period prevalences were 2.4% and 2.0%, respectively. Cumulative prevalence of asthma was significantly higher in Indians (6.6%) and Malays (6.0%) than in Chinese (3.0%); period prevalences of asthma were 4.5% in Indians, 3.3% in Malays, and 0.9% in Chinese. Ownership of cats or dogs was more frequent in Malays (15.4%) and Indians (11.2%) than in Chinese (8.8%). Rugs and carpets were also more frequently used by Malays (52.2%) and Indians (40.7%) than by Chinese (8.9%). Current smoking prevalences were higher in Malays (27.3%) than in Indians (19.4%) and Chinese (23.0%). Malays and Indians did not have higher rates of atopy (11.1% and 15.2%, respectively) than Chinese (15.4%). Adjustment for these factors in multivariate analyses reduced the greater odds of asthma in Malays and Indians, but not to a significant extent. CONCLUSIONS--There are ethnic differences in the prevalence of asthma in Singapore which are not entirely explained by differences in smoking, atopy, or other risk factors. Other unmeasured environmental factors or genetic influences are likely to account for residual differences in the prevalence of asthma.