The prognosis of early childhood respiratory illness has been studied by analysis of historic cohorts using data recorded since 1967 in the continuous morbidity registration project of the University of Nijmegen. The characteristics of this project are discussed, including the stability of the practice (study) population, a factor which allows longitudinal analysis to be performed.The development of respiratory morbidity during the first eight years of life has been analysed and a group of 710 patients characterized on the basis of their respiratory morbidity in the first two years of life. In general the highest levels of respiratory morbidity are seen in the first years of life, with a gradual decline over the years studied. Upper respiratory tract infections are largely responsible for this morbidity. An elevated level of episodes of respiratory illness in patients aged two to seven years was observed when the following characteristics of early life respiratory morbidity were present: more than three episodes of respiratory illness, lower respiratory tract infections or catarrhal conditions. However, by the age of seven years the level of morbidity of the children with these characteristics was very similar to that of children without these characteristics. Respiratory morbidity in childhood is predominantly benign and self-limiting. Only children with atopic conditions in early childhood showed a higher prevalance of asthma and chronic bronchitis at the end of the study period.