To investigate potential risk factors for the development of childhood asthma, the authors undertook a longitudinal study using a cohort of 770 children aged 5-9 years from East Boston, Massachusetts, that has been under study since 1975. The disease outcome considered was age at first onset of asthma, as determined by parental or self-reporting of a physician's diagnosis. Potential risk factors were evaluated specifically in relation to their presence antecedent to a diagnosis of asthma. Standardized questionnaires were used to obtain childhood illness histories, environmental exposures, and the asthmatic and atopic statuses of first-degree relatives. Ninety-one cases of asthma were identified from 1975 to 1988 (57 males and 34 females). Significant sex-adjusted relative risk estimates were seen for antecedent pneumonia, bronchitis, hay fever, sinusitis, parental asthma, and parental atopy. Neither bronchiolitis, eczema, croup, personal cigarette smoking, maternal smoking, paternal smoking, nor delivery complications bore an apparent relation to the development of asthma. A history of parental asthma or parental atopy did not significantly alter the sex-adjusted relative risk estimates for pneumonia, bronchitis, hay fever, or sinusitis. These results support the hypothesis that asthma is a multifactor disease whose expression is dependent on both familial and environmental influences.