Postnatal growth of the lung has been the subject of investigation and controversy for many years. Koelliker (1881) believed that the adult structure of the lung was already present in the newborn infant and growth involved expansion only. This view was supported in part by Short (1950), who considered that lung growth was determined by purely physical factors, and thought that simple distension of the newborn rabbit lung resulted in a structure resembling that of an adult lung. More recent work has indicated that the lungs of newborn animals are qualitatively different from those of adults of the same species in that alveoli are few or absent at birth. Thus most or all alveoli are formed after birth in rats (Weibel, 1967; Burri, Dbaly & Weibel, 1974a), rabbits (Engel, 1953), cats (Engel, 1953; Dingier, 1958), andman (Dubreuil, Lacoste & Raymond, 1936; Boyden & Tompsett, 1965; Reid, 1967). The way that alveoli are formed is controversial. Subdivision of peripheral lung units, alveolarization of non-alveolated airways (bronchioles), peripheral airway branching and alveolar'sprouts' have all been described. Boyden & Tompsett (1961, 1965) have stressed the fact that the peripheralairways are different in the newborn dog and human from adults of thesame species. They have termed the relatively large, simple units in newborn puppies' terminal saccules'. Similar observations have been made by Weibel (1967) and, in his laboratory, Burni and his colleagues (1974a, b) have recently made a systematic survey of the postnatal growth of the lung in the rat. They demonstrated that the neonatal rat has no true alveoli, and that respiratory exchange is carried out in large smooth-walled sacs ('primary saccules'). The walls of these saccules were called'primary septa'. Primary saccules differed from alveoli by their large size and lack of complexity, and primary septa from nature inter-alveolar septa by the presence of a double capillary system in the former. They thought that the primary saccules were subdivided in the postnatal period by slender crests (termed'secondary crests or septa') into alveoli. Secondary septa consisted of a covering of Type I alveolar epithelium and a core of interstitial cells, capillaries and connective tissue. Subsequent development of the lung was marked by a decrease in interstitial tissue and an increase in lung complexity. The purpose of this paper is to amplify these observations concerning postnatal growth, using the scanning electron microscope, to provide data concerning lung