We have developed a quantitative model to study the impact of screening for the hepatitis B surface antigen using either second-generation (counterimmunoelectrophoresis) or third-generation (radioimmunoassay) tests. This model was fashioned for general use but was applied to data on hepatitis for India, in an attempt to determine what factors would most influence the decision to institute a screening program for the surface antigen in that country. Screening for the surface antigen should reduce post-transfusion hepatitis B significantly--second-generation testing by 81% and third-generation testing by 99.4%. The marginal cost of substituting counter-immunoelectrophoresis testing for no testing is 557 rupees per case averted, and that of substituting radioimmunoassay for counterimmunoelectrophoresis testing is 1,120 rupees per case averted. Sensitivity analyses suggest that the major impact on costs will result not from changes in the sensitivity or specificity of new screening tests, but rather in their costs.