We report the results of segregation analyses for wheeze before and after a history of respiratory allergy was taken into consideration. The analyses were based on data from 309 nuclear families with 1,053 individuals living in the town of Humboldt, Saskatchewan in 1993, and were performed by using the REGD program of the SAGE package. For adults, information on wheeze and history of respiratory allergy was provided by themselves, and for children, by their parents. Segregation analyses were first conducted before adjustment for history of respiratory allergy. Other covariates were adjusted including sex, current smoking, household exposure to tobacco smoke, and type of house. A single locus model with residual familial effects fit the data well, but none of the Mendelian models (recessive, dominant, and codominant) could be distinguished. The no-parent-offspring-transmission hypothesis was rejected. However, when the variable of respiratory allergy was included in the models as a covariate, both Mendelian and environmental hypotheses were rejected. The Mendelian model had a relatively lower value of Akaike's Information Criterion than did the environmental model (1095.56 versus 1111.24). The data suggest that a single locus gene explains a portion of wheeze that is related to respiratory allergy, and that common environmental factors and/or polygenes also account for a certain familial aggregation of wheeze.