Murine recombinant soluble ICAM-1 (sICAM-1) was immobilized on polystyrene microspheres. The binding of sICAM-1-coated microspheres to splenic T cells required previous activation of the cells and was inhibited by antibody to LFA-1 (CD11a/CD18), indicating that sICAM-1-coated microspheres bind exclusively to high-avidity LFA-1. The cytoskeleton inhibitor cytochalasin B did not inhibit the binding of sICAM-1-coated microspheres to PMA-activated splenic T cells, whereas their adhesion to sICAM-1 immobilized on microtiter wells was almost completely inhibited. The murine T hybridoma T28 cells on activation with PMA also bound sICAM-1-coated microspheres, and the binding sites on the cell surface seemed localized on some of the cells, whereas fluorescence staining showed an even distribution of LFA-1 on the cell surface. In contrast, the murine B cell line A20A8 and monocytic line P388 showed a more even distribution of sICAM-1 binding sites. To further investigate the distribution of high-avidity LFA-1, murine fibroblast L cells expressing LFA-1 were generated by gene transfer. The transfected L cells constitutively expressed high-avidity LFA-1 and bound sICAM-1-coated microspheres without previous activation. Interestingly, the binding sites seemed highly localized on most cells. In contrast, the binding sites for anti-LFA-1 Ab-coated microspheres were randomly distributed on the transfected L cells. Furthermore, fluorescence staining also revealed a uniform punctate distribution of LFA-1 on the surfaces of these cells. These results show that 1) sICAM-1-coated microspheres represent a useful tool in identifying high-avidity LFA-1, 2) the binding of sICAM-1-coated microspheres to high-avidity LFA-1 does not require an intact cytoskeleton, and 3) the cell surface distribution of high-avidity LFA-1 can be different from that of LFA-1 in general, and the former seems highly localized on some cells.