The 1970 National Health Interview Survey included questions on respiratory disease and smoking habits. The new data were released in July 1974. Data consisted of information on approximately 116,000 persons from 37,000 households selected randomly from 357 primary sampling units. To test a hypothesis about familial clustering of chronic respiratory disease in households, we selected as index households those having an adult (aged 35-54 years) reporting a diagnosis of asthma, bronchitis, or emphysema and also having first order relatiaves less than age 35. Index households were matched with households from the same neighbourhood having an adult aged 35-54, of the same sex as the diseased person in the index household, without disease and with all other adults 35-54 without disease, and which had first order relatives less than 35 living in the same household. Analysis was carried out using Cochran's d-test to compare frequency of respiratory disease in persons less than 35 in each group. There was a strong association (P less than .001) between persons over 35 with chronic respiratory diseases and the disease rate in their first order relatives. The association could not be explained by differences in demographic variables and smoking habits.