Background: There is no information whether galectin-9 (a novel eosinophil chemoattractant) was associated with pathogenesis of eosinophilic disorders. Methods: We assessed the expression of galectin-9 with imunostaining and in situ hybridization both in the lesion of angiolymphoid hyperplasia with eosinophilia, and peripheral blood eosinophils of eosinophilic patients (E-Eos) in comparison with those of normal volunteers (N-Eos). Regulation of expression of galectin-9 on eosinophils and the effect of galectin-9 on apoptosis of eosinophil were also evaluated. Results: Many eosinophils infiltrating the site were positive for galectin-9. Surface and intracellular immunoreactive galectin-9 was more evident in E-Eos than N-Eos. When eosinophils were cultured with IL-5 in vitro, the surface galectin-9 expression of E-Eos was significantly downregulated, although that of N-Eos was not affected. Treatment of eosinophils with dexamethasone or anti-Fas antibody significantly upregulated the surface galectin-9 expression of E-Eos. In contrast, dexamethasone partially downregulated the surface galectin-9 of N-Eos, although anti-Fas antibody failed to affect on the surface galectin-9 expression. We also found that recombinant galectin-9 significantly suppressed apoptosis of E-Eos (p = 0.0431), whereas it apparently enhanced apoptosis of N-Eos (p = 0.0173). Furthermore, dexamethasone-induced apoptosis of N-Eos was significantly suppressed by galectin-9 (p = 0.0431), whereas galectin-9 failed to induce significant change in dexamethasone-induced apoptosis of E-Eos. In contrast, apoptosis induced by anti-Fas antibody in both N-Eos (p = 0.0431) and E-Eos (p = 0.0431) was enhanced by galectin-9. Conclusions: These findings suggested that galectin-9 was produced by eosinophils, and galectin-9 showed heterogeneous effects and kinetics to eosinophils, and this factor might be one of crucial factors in eosinophilic inflammation.