The acute inflammatory process is characterized by an orderly progression of events; an initial phase of early neutrophil accumulation and a later phase of mononuclear cell (including monocyte) accumulation. The mechanisms which control the transition from one phase to the other are largely unknown. We present a rabbit model of C5 fragment (C5f)-induced lung inflammation in which purified radiolabeled peripheral blood neutrophils and monocytes were used as probes to monitor the retention and emigration of these leukocytes into well localized areas of inflammation. Neutrophil preparations (greater than 95% pure) were isolated by discontinuous plasma-Percoll density gradients, and monocyte preparations (greater than 91% pure) were isolated by counterflow cell elutriation, labeled with 111Indium-tropolonate, and intravenously infused into separate recipient animals. The monocytes circulated with a half-life of approximately 30 hours. The retention of labeled monocytes or neutrophils within the lung was monitored scintigraphically. C5f-induced monocyte lung retention was delayed 2 to 4 hours compared with neutrophil lung retention. Radiolabeled neutrophils were selectively retained in the area of C5f-induced inflammation (right cranial lung lobe, RCL) as early as 20 minutes after the induction of the inflammatory response, reached a maximum by 2 hours, and were not retained by 48 hours after C5f instillation. The signal inducing C5f-induced monocyte lung retention was shown to be transient. Monocytes were selectively retained in the RCL if the area of inflammation was induced 2 to 4 hours but not 15 minutes or 16 hours before their infusion. The time course of C5f-induced monocyte migration into the alveolar space determined by lavage analysis was delayed 2 to 3 hours compared with neutrophil migration. Neutrophils selectively migrated into the RCL 1 to 2 hours after the induction of the inflammatory response, reached a maximum by 4 hours, and had disappeared by 48 hours. Radiolabeled monocytes selectively migrated into the RCL 3 to 4 hours after the induction of the inflammatory response, reached a maximum by 4 hours, and remained present through 48 hours. The total number of labeled and unlabeled mononuclear cells present in the C5f-treated RCL lavage at 48 hours was significantly increased above controls. The signal for this monocyte migration (as for lung retention) was shown to be transient in that radiolabeled monocytes did not migrate when infused 16 hours after the induction of the inflammatory response. C5f did not induce monocyte lung retention nor monocyte migration into the alveolar space of animals rendered neutropenic.(ABSTRACT TRUNCATED AT 400 WORDS)