This study was conducted to investigate the serum and urine levels of survivin in patients with breast cancer and the relationships with known prognostic parameters and therapy. Forty-three patients with breast cancer and 21 healthy control subjects were investigated. Serum samples were obtained on the first admission before adjuvant and metastatic treatment were given and after two cycles of chemotherapy. Serum and urine survivin levels were determined using enzyme immunometric assay (EIA) and enzyme-linked immunosorbent assay (ELISA), respectively. There was no significant difference in the baseline serum and urine levels between patients with breast carcinoma and healthy controls (p = 0.19 and p = 0.84, respectively). None of the prognostic parameters analyzed were significantly correlated with the urine survivin concentrations. This was also true for serum survivin values, except for nodal involvement. Serum survivin levels were significantly higher in the patients with nodal involvement compared with node negatives (p = 0.043). However, serum survivin levels were not influenced by the number of involved nodes (p = 0.77). No significant correlation was found between the serum and urine levels of survivin (r = 0.15, p = 0.27). Serum and urine levels did not change significantly after chemotherapy (p = 0.59 and p = 0.50, respectively). In conclusion, the result of this study suggested that serum survivin level could be a sensitive marker for detecting metastases in lymph nodes from breast cancer patients. However, much research continues in this field, and exciting new knowledge will ultimately emerge.