The distribution of immune complexes has been studied in mouse spleen stimulated to contain many germinal centers (GC's). Horseradish peroxidase (HRP)-anti-HRP complexes were used as an appropriately precise and sensitive model. We were primarily interested in the relative abilities of three cell types to interact with complexes: lymphocytes, macrophages, and follicular dendritic cells (FDC's). The latter are distinctive, nonendocytic, stellate cells located primarily at the transition of mantle and GC zones of 2 ~ lymphoid follicles (Chen, L. L., J. C. Adams, and R. M. Steinman, 1978, J. Cell Biol. 77"148).Binding of immune complexes to lymphocytes could not be visualized in situ. Macrophages avidly interiorized complexes into lysosomes, but did not retain them extracellularly. In contrast, FDC's could retain HRP-anti-HRP extracellulady under appropriate conditions, but did not endocytose them. Cytochemical reactivity accumulated progressively on FDC's 1-6 h after administration of complexes i.v., remained stable in amount and location for 1 day, and then was progressively lost over a 1-to 5-day period.Several variables in the association of complexes with macrophages and FDC's were pursued. Only 1 /~g of complexed HRP had to be administered to visualize binding to both cell types. Macrophages interiorized complexes formed in a wide range of HRP/anti-HRP ratios, while FDC's associated with complexes formed in HRP excess only. Quantitative studies with [12SI]HRP-anti-HRP demonstrated that 20% of the splenic load of HRP associated with FDC's. Complexes formed with an F(ab')~ anti-HRP were distributed primarily in macrophages. When the levels of the third component of serum complement were depleted by prior treatment with cobra venom factor, uptake of complexes by macrophages was reduced some 50% whereas association with FDC's was abolished.The fact that antigen excess complexes are retained extracellularly strengthens the idea that they are immunogenic. Finally, the association of complexes with FDC's seems to retard the entry of antigen into the GC proper.J. CELL BIOLOGY 9 The Rockefeller University Press 9 0021-9525/78