STUDY OBJECTIVE--The aim was to investigate predictors of childhood lower respiratory tract illness in two generations, and predictors of adult lower respiratory disorders in the first generation. DESIGN--Data on respiratory health and environmental factors from a national birth cohort study were examined from birth to 36 years. Data were also collected on the parents of the subjects and on the subjects' first born offspring from birth to eight years. Main outcome measures were: reports of lower respiratory tract illness before 2 years; lower respiratory tract illness of a week or more between age 20 and 36 years; regular phlegm production at 25 and 36 years; reports of wheeze or asthma at age 36 years; peak expiratory flow rate (PEFR) at age 36 years measured by nurses during home visits; and mothers' reports of lower respiratory illness in first born offspring before 2 years. SUBJECTS--Subjects were a sample of 5362 single, legitimate births taken from all those occurring in England, Wales, and Scotland in one week in 1946, and studied regularly from birth to age 43 years. Data on the subjects' parents and on their 1676 first offspring born while they were aged 19-25 years were also collected. MAIN RESULTS--Lower respiratory tract illness before 2 years fell from 25% in the population born in 1946 to 13% in their first born offspring. In those born in 1946, poor home environment, parental bronchitis, and atmospheric pollution were the best predictors of lower respiratory illness before 2 years, and these three factors and childhood lower respiratory illness and later smoking were the best predictors of adult lower respiratory tract problems. Risk factors for lower respiratory illness in the offspring were manual social class, parental and grandparental lower respiratory disease, and parental smoking. CONCLUSIONS--Risks for adult lower respiratory problems accumulated in childhood through illness, poor social circumstances, and atmospheric pollution. Smoking exacerbated early life risks and was an independent risk factor. In the offspring generation, parental smoking was a risk factor for early life chest illness, together with parental illness and low social class. Reduction of prevalence in the offspring generation was probably accounted for by improvement in home circumstances, reduced atmospheric pollution, and lower rates of parental lower respiratory illness, but current rates of smoking seem likely to prevent much further reduction in early life lower respiratory illness, and thus in this aspect of risk for subsequent adult lower respiratory problems. The accumulation of risk in childhood and adolescence for later adult problems implies a long time scale for the reduction of adult lower respiratory disorders.