Immunoglobulin allotyping in 925 cases of disputed paternity provided evidence of exclusion for 70 alleged fathers. Combining results from erythrocytic antigen, enzyme, and serum protein tests, 230 men were proven to be falsely accused; in 15 cases the immunoglobulin allotype provided the only evidence for exclusion, in 67 no exclusion would have been identified if testing had been limited to ABO, Rh, and MNSs. Various cases in this study illustrate the importance of using an extensive battery of immunoglobulin reagents in order maximally to exclude falsely accused men, and to identify infrequent haplotypes, which might erroneously be interpreted as indirect exclusions or may indicate a high likelihood of paternity. The problems of Gm typing of very young children are described. When child is less than 6 months of age, direct exclusions may be missed by the Gm allotypes; indirect exclusions are valid only when the phenotype differs from that of the mother. Km allotypes are not age-dependent.