The detection of TSH-receptor antibodies (TRAb) in patients with Graves' disease is routinely used in nuclear medicine laboratories. This determination has been possible for approximately 3 years with a second generation human TRAb assay. Studies showed that this TRAb determination is diagnostically more sensitive compared to established, porcine TRAb assays. Objective: The aim of our study was to investigate, based on a ROC analysis, whether TRAb determination with the new, second generation assay allows a dependable statement about probability of occurence of relapse after radioiodine therapy in patient suffering from Graves' disease. Methods: 57 patients were examined with the DYNOtest® TRAKhuman (BRAHMS Diagnostica AG, Hennigsdorf) directly before and six months after therapy with radioiodine (dose: 150 Gy). A ROC-analysis was performed to determine positive/negative predictive values depending on different cut-off values. Results: Whereas 21/57 patients became eu- or hypothyroid after six months, 36/57 patients relapsed. Nonrelapsed patients showed a significant lower median TRAb titer (4.2 IU/I vs. 19.2 IU/I; p <0.05) compared to relapsed patients. But the positive predictive value conducted 63 and 66, 62 and 66 as well as 63 and 69% (before and after therapy) linked with the cut-offs 1.0, 1.5, and 2.0 IU/I. So it was in areas also achieved by the first generation porcine radioreceptorassay. Conclusion: An increased sensitivity is achieved undoubtedly with the new DYNOtest® TRAKhuman in the diagnostic of Graves' disease. It is not held over the established radioreceptorassay concerning the positive predictive value for relapsing patients.