Background: The value of CA15-3(cancer antigen 15-3) marker in early detection of breast cancer recurrence has been studied in several prospective trials. But the results of these studies are different. This may be due to variable cutoff points used for analysis, different intervals between CA15-3 measurements and the differences between patients population. This study was done to examine the predictive value, sensitivity and specificity of `CA15-3 in detection of breast cancer recurrence in Iranian patients, using a commercial available ELISA assay with a cut-off value of 30 u/ml. Methods: For this purpose serial blood samples and clinical data of 133 female patients with breast cancer referring for adjuvant treatment to Radiation Oncology Department of Shiraz University of Medical Sciences from 1379 to 1382 were collected and for all cases of clinically suspected recurrence, routine methods were used for documentation. Results: Mean age of patients was 45.6 years (SD 11) and mean time of follow up was 17.5 months. Recurrence was documented in 39 patients who were classified into 4 groups: patients with documented recurrence and early elevatedmarker (true positive,32 patients), patients with documented recurrence but no elevated marker(false negative, 7 patients), those without recurrence and normal marker (true negative,76 patients) and those without recurrence but elevated marker (false positive,18 patients). According to these data sensitivity, specificity and positive and negative predictive values of CA15-3 were 82,80.85,64 and 91.5 percent respectively. Conclusion: CA15-3 with cut-off point of 30 u/ml has an adequate sensitivity and specificity for early detection of breast cancer recurrence in Iranian patients.