A sensitive sandwich enzyme immunoassay for human pulmonary surfactant protein D (SP-D) was developed and used to examine the blood SP-D levels of drowning victims. Human SP-D was purified from amniotic fluid by chromatographic methods, and an antibody against human SP-D was prepared. A polystyrene ball coated with anti-SP-D IgG was incubated with purified human SP-D, and then with anti-SP-D Fab'-peroxidase conjugate. Peroxidase activity bound to the polystyrene ball was assayed by fluorometry using 3-(4-hydroxyphenyl)propionic acid as the hydrogen donor. The detection limit of human SP-D was 5.2 pg per assay tube. Examination of cross-reactions of this sandwich enzyme immunoassay with proteins from other human organs showed it to be highly specific for lung, and Northern blot analysis detected specific SP-D mRNA expression only in lung. The SP-D concentration of normal human serum was 6.4+/-2.7 (mean+/-S.D.) ng ml(-1) (n=20). The recovery rates of 0.52 ng and 5.2 ng SP-D added to 5 microl normal human serum were 93.6+/-2.7% and 93.6+/-6.1%, respectively. Blood SP-D levels of victims from the saltwater drowning group (n=14) revealed higher concentrations (105.8+/-53.7 ng ml(-1)), while freshwater drowning victims (n=12) were estimated to be 74.1+/-43.9 ng ml(-1). The SP-D levels of 15 subjects who died of hemorrhage (n=5), heart failure (n=8), traumatic shock (n=1), and electrocution (n=1) were lower (22.0+/-8.5 ng ml(-1)), and those of asphyxia victims (n=10) were slightly higher (36.2+/-17.1 ng ml(-1)) than those of other causes of death, except for drowning. These results suggest that in drowning victims, SP-D flowed into the systemic circulation by physiological and physical mechanisms, and the differences of blood SP-D levels between saltwater drowning and freshwater drowning victims are presumed to be influenced by the type of agony and/or the length of survival time in water.