An important criterion for the clinical use of a new imaging technique is the correct reproducibility of interpretation. Forty-six paired immunoscintigraphic examinations were performed on 43 patients with suspected ovarian carcinoma recurrence using F(ab')2 fragments of OC125 antibody labeled first with indium-111 and then with iodine-131. Planar scintigraphy (PS) and emission computed tomography (ECT) images were interpreted blindly and separately by three observers, and reproducibility was evaluated by a kappa concordance index. Intra- and interobserver reproducibility were generally satisfactory (kappa values of 0.6 and 0.7, respectively). Binomial analysis of kappa values for ECT showed the superiority of indium-111 for intraobserver (p = 0.035) and interobserver (p = 0.0039) study. However, for PS there was no significant difference in reproducibility with the two radionuclides.