We have attempted to determine whether interleukin-5 (IL-5), a cytokine that selectively affects eosinophil (as opposed to neutrophil) differentiation and activation, also modulates eosinophil migrational responses. Using a modified Boyden chemotaxis assay, IL-5, IL-3, and granulocyte-macrophage colony-stimulating factor (GM-CSF) gave a weak locomotory response for eosinophils from normal nonatopic subjects (optimal at 10(-11), 10(-8), and 10(-9) mol/L, respectively), but not for eosinophils from subjects with an eosinophilia associated with asthma and/or allergic rhinitis. In contrast, IL-5 and IL-3 had no effect on neutrophils, while GM-CSF was chemotactic for neutrophils over a limited concentration range, optimal at 10(-8) mol/L. When eosinophils from normal subjects were incubated with IL-5 (10(-9) mol/L), the locomotory response to platelet-activating factor (PAF; 10(- 8) mol/L, P less than .05), leukotriene B4 (LTB4; 10(-6) mol/L, P less than .01), and N-formyl-methionyl-leucyl-phenylalanine (FMLP; 10(-8) mol/L, P less than .01) was significantly enhanced. The percentage enhancement of eosinophil locomotion by IL-5 was greater for eosinophils from normal as compared with subjects with an eosinophilia associated with asthma (P less than .05 for PAF and LTB4; P less than .01 for FMLP). Preincubation of eosinophils from normal subjects with IL-5 (10(-9) mol/L) attenuated the subsequent locomotory response to IL- 5 (10(-12) and 10(-11) mol/L, P less than .05). Therefore, the observed refractoriness of eosinophils from eosinophilic subjects to both directional migratory and priming effects of IL-5 in vitro, may reflect a deactivation process resulting from prior exposure in vivo. The selective priming of eosinophil but not neutrophil locomotion by IL-5 suggests that this cytokine may play a significant role in the preferential accumulation of eosinophils at sites of allergic inflammation.