To diagnose the typhoid fever rapidly and accurately in clinically suspected patients, the levels of IgG subclass antibody were measured by enzyme-linked immunosorbent assay(ELISA). With symptom, blood culture and agglutination test, tested persons were categorized into 6 groups as typhoid fever, FUO, paratyphi A or B, other bacterial infctions, cancers, and control. ELISA was performed on the polyvinyl chloride plates coated with killed whole cell($10^8;cell/ml$) of S. typhi 0901W by poly-L-lysine applied as binding substance (and polyvinyl chloride as solid phase). The distribution of the level of IgG subclass antibodies in each group was analyzed and compared with other groups. The results obtained were summarized as follow: 1. The optimal dilution of the sera from patients with typhoid fever was 1:160, and those of the sheep anti-human IgG subclass and the peroxidase conjugated rabbit anti-sheep IgG were 1:4000 and 1:5000, respectively. 2. The absorbance levels of IgG subclass in the sera of typhoid fever patients were as follows; a) IgG1 value is $0.439{pm}0.110$ b) IgG2 value is $0.416{pm}0.165$ c) IgG3 value is $0.449{pm}0.145$ d) IgG4 value is $0.525{pm}0.154$ IgG subclass levels in the sera of typhoid patients were much higher than in control group and patient with paratyphi A or B as well as other infectious diseases. The sensitivity and the specificity in differential diagnosis of typhoid fever and other febrile diseases were 92% and 79% in the assay of IgG1 respectively, whereas those in the assay of IgG2 were 97% and 72%, respectively (above absorbance 0.3). 3. The absorbance levels of IgG subclass in the serial sera of typhiod fever patients tend to decrease to the level of absorbance 0.3 in 10 months from the onset of illness. 4. The order of absorbance levels of IgG subclass in the serum of each group were typhoid fever, paratyphi A or B, other infectious diseases, control and cancer. 5. For the serodiagnosis of typhoid fever against other febrile diseases, the sensitivity and the specificity in the assay of IgG2 activity were 76% and 93% in absorbance 0.4, respectively. 6. In the distribution of the level of each IgG subclass in the sera of FUO patients which were suspected of typhoid fever, the positive rate was ranged from 36% to 82%. This suggest that more than 50% of FUO patients are caused by S. typhi.