The subchronic effects of selenomethionine (SeM) and sodium selenite (SeL) on several immunologic, hematologic, and serologic parameters in mallards were measured, using concentrations in drinking water of 0, 0.5, and 3.5 mg/L selenium (se) as SeL and 2.2 mg/L (Se) as SeM. Cyclophosphamide (CP) was used as an immunosuppressive control at 20 mg/L. A battery of in vivo and in vitro immunologic assays was performed on each bird throughout the 12 week study. The SeM-treated group also exhibited a significantly increased serum alanine aminotransferase (ALT) activity and an increased plasma glutathione peroxidase (GPX) activity (p < or = 0.05). Body weight and water consumption of treated birds did not differ from controls. Organ weights were not significanlty affected by any Se treatment. Sodium selenite-treated birds displayed not detectable differences in immune-function or Se accumulation in tissues as compared to controls. Serum ALT activity was significantly increased in the 3.5 mg/L group, although to a lesser extent than in SeM-treated birds. Cyclophosphamide significantly depressed white blood cell number, testes weights, and also suppressed the DTH reaction. Concentrations of Se as SeL did not affect the immune system, whereas low concentration of SeM (2.2 mg/L Se) appeared to suppress certain aspects of the mallard immune response.