This report gives a quantitative description of the kinetics of the pulmonary macrophages and their direct precursors during the acute inflammatory reaction in the lungs induced by intravenous injection of heat-killed bacillus Calmette-Guérin (BCG) into specific-pathogen-free mice. After BCG injection, the total number of pulmonary macrophages isolated by lavage and subsequent enzyme digestion of lung tissue increased to 225% of normal within 12 h and, after a minor decrease, rose to a maximum of 250% of normal at 96 h, followed by a decrease to 150% at 144 h, the end of the observation period. The number of circulating monocytes doubled in the first 48 h and stayed close to that level. In vivo and in vitro labeling with [3H]-thymidine showed that an influx of monocytes transforming into pulmonary macrophages was mainly responsible for the population increase. A temporary increase in the number of locally dividing pulmonary macrophages--manifested by an increased in vitro labeling index, reaching a maximum of 9.6% 72 h after BCG injection--made a minor contribution to the population increase. All pulmonary macrophages were classified according to morphological criteria as alveolar-macrophage-like (AML) or non-alveolar-macrophage-like (NAML), and their respective characteristics were established. The in vivo labeling data showed NAML to represent exudate macrophages derived from circulating monocytes entering the interstitial tissue, and these cells changed morphologically into AML upon entering the alveolar hypophase. This mechanism was confirmed by the finding that the interstitially deposited BCG were found first inside NAML and later in AML. The in vivo labeling data showed that local production was mainly a result of division of macrophages that were morphologically identical with normal alveolar macrophages. The former cells, however, derived most probably recently from the circulation, because the turnover of the total population was very high before local macrophage production became maximal. In mice treated with HC before the injection of BCG, this population increase was absent, because of virtual abolition of the initial monocyte influx and absence of the increased local production of macrophages. Calculations showed that the monocyte influx in the first 48 h amounted to approximately 4 x 10(6) cells, i.e., eight times that found in the normal steady state, and that the efflux of pulmonary macrophages in that period amounted to approximately 3.5 x 10(6) cells, i.e., seven times the normal efflux. The local production over the total period of 144 h was only three times that found normally. The results of these quantitative studies show that the increase of the pulmonary macrophage population during an acute inflammation is brought about mainly by monocyte influx and to a minor extent by a temporary increased local production of macrophages. Disposal of interstitially deposited BCG occurred by phagocytosis by local macrophages and the subsequent efflux of the latter.