A B S T R A C T Therapeutic doses of corticosteroids frequently induce eosinopenia; however, the mechanism(s) involved remain obscure. To investigate this question, we studied the effects of corticosteroids on eosinophil adherence and migration. Eosinophils from normal donors were prepared by dextran sedimentation and Hypaque gradient centrifugation to 45-96% purity. Adherence was measured by filtration of whole blood and isolated eosinophils through nylon wool columns. Before prednisone administration, adherence was 83.8+3.2% for eosinophils in heparinized blood and 82.1±3.2% for isolated eosinophils. 4 h after oral prednisone administration whole blood eosinophil adherence was reduced to 53.9+10.7%; at 24 and 48 h adherence was normal. In contrast, isolated eosinophils showed no decrease in adherence 4, 24, or 48 h after corticosteroid administration. Similarly, in vitro addition of hydrocortisone to isolated eosinophils at 0.01 and 2.0 mg/ml did not reduce adherence. Eosinophil migration was tested in modified Boyden chambers by "lower-surface" and "leading-front" methods, using zymosan-activated serum and buffered saline to assess chemotactic and random migration, respectively. In vitro incubation of eosinophils with hydrocortisone or methylprednisolone produced a dose-dependent inhibition ofchemotaxis. Using lowersurface methods the minimal concentration effecting substantial inhibition was 0.01 mg/ml for both drugs. At 2.0 mglml hydrocortisone and methylprednisolone inhibited eosinophil chemotaxis 82.6±4.4% and 85.0±3.5%, respectively. Using leadiig-front chemotaxis techniques significant inhibition was detected at 0.001 mg/ml hydrocortisone. Eosinophils incubated and washed free of corticosteroids responded normally to chemoattractants, indicating that the inhibitory effect of these drugs was reversible. HydroAddress reprint requests to Dr. Altman. Received for publication 21 March 1979 and in revised form 18 September 1980. 28 cortisone at 2 mg/ml inhibited random eosinophil migration, although this effect was not apparent at lower concentrations. Corticosteroids did not act as chemotactic factor inactivators and were not toxic as measured by trypan blue exclusion. Eosinophils obtained from donors who had received 40 mg of prednisone orally for four days showed normal chemotactic responses, probably reflecting the fact that the cells were washed free of plasma before testing. In contrast, incubation of eosinophils in plasma from donors who had received a 300-mg bolus of hydrocortisone induced 46.1±4.5% more inhibition of chemotaxis than did incubation in normal plasma. These results indicate that: (a) eosinophil adherence is transiently reduced following in vivo corticosteroid administration, (b) eosinophil chemotaxis is inhibited by both in vitro and in vivo administration of corticosteroids, and (c) the chemotaxis inhibiting effect is nontoxic, cell-directed, dose-dependent and reversible. Inhibition of eosinophil adherence and chemotaxis may in part explain how corticosteroids produce eosinopenia a...