Pulmonary aspergillosis usually develops on the basis of systemic immunosuppression and/or local impairments of respiratory system. Diagnosis of pulmonary aspergillosis has many difficulties. Chest X-ray findings of most cases are complicated with pre-existing changes due to the underlying diseases, and the detection rate of the pathogenic fungi from clinical specimens is unsatisfactorily low. Therefore, immunological or serological diagnosis is urgently required and precipitation-in-gel method has been widely applied. In this report, we compared clinical usefulness of the determination of anti-aspergillus antibodies by ELISA with that of precipitation-in-gel method. ELISA was carried out according to the method previously reported by us (Yamamoto S. et al.: Kekkaku 62: 549, 1987). About two-thirds of 45 healthy adults (control) did not show any detectable IgG anti-aspergillus antibody and mean of IgG anti-aspergillus antibody titer of the control group was 28.97. Patients, who had shown positive culture of fungus or was clinically diagnosed or strongly suspected as pulmonary aspergillosis, showed significantly high anti-aspergillus IgG antibody titer in comparison with the control group. Further, patients who were positive in precipitation-in-gel tests showed significantly higher IgG antibody titers than those who were negative in that test. IgG antibody titer determined by ELISA corresponded with clinical diagnosis much more exactly than the results of precipitation-in-gel test. Further, the results obtained by ELISA were objective and quantitative in comparison with the latter test. We concluded that ELISA was much superior to precipitation-in-gel test and that ELISA IgG antibody titers 2500 or more were confirmative and those between 570 and 2500 were strongly suggestive for the diagnosis of aspergillosis. IgM anti-aspergillus antibody titers were not different among healthy control group and patient groups, and could not be used for the diagnosis.