The occlusion technique to measure total respiratory system compliance (Cocc) was used in 28 sedated infants with a variety of cardiopulmonary diseases and 14 anaesthetised infants during the first 2 years of life. In this report, we describe some of the potential problems in the technique and how to avoid them. Invalidation of the occlusion technique because of consistent failure to relax during brief airway occlusions occurred in only four of 42 infants despite the fact that only four infants were studied during the first month of life and 17 were between 12 and 26 months old. The technique was invalidated in two intubated infants because of persistent leaks around the endotracheal tube. In the remaining 36 infants, data sometimes had to be excluded as a result of instability and end-expiratory volume, intermittent leaks, or failure to relax during occlusions performed at low lung volumes or during inspiration. However, by performing 15-25 occlusions per infant, it was possible to obtain sufficient reliable data for accurate analysis of Cocc in all these infants. Providing that the potential errors and limitations of the occlusion technique are recognized, it appears to be applicable to a wide range of healthy and sick infants.