Background: Neoadjuvant chemotherapy (NAC) with epirubicin/cyclophosphamid (EC) followed by docetaxel (D) is currently a standard of care therapy in women with early, high-risk breast cancer (BC). New approaches aim to improve the outcome by combining chemo- with immunotherapy. It is therefore of great interest if chemotherapeutics differ in their effect on the immune system and if some substances are superior combination partners than others. Methods: 79 BC patients, who participated in the ABCSG-34 trial, were included. 39 patients were treated with 6 cycles of EC followed by 6 cycles of D and 40 received the reverse sequence (D→EC). Blood was collected before and after 6 cycles. The plasma levels of a variety of immune mediators were determined by multiplex bead array assay. The response to therapy was measured by Residual Cancer Burden (RCB)-score. A score of ≤1.36 was determined as good response. Lymphocyte activation was assessed after stimulation with phytohaemagglutinin (PHA) by flow cytometric analysis of IFNgamma. Additionally, lymphocytes of 6 healthy probands were stimulated with PHA and treated with E or D. The stimulation was quantified by measuring cluster formation after 5 days. Further, a human BC cell line (SK-BR3) was treated with E or D. The induced cell death was determined morphologically as well as by flow cytometry after staining of phosphatidylserine, Sub-G1 DNA and active caspase-3. Results: The treatment of BC patients with 6 cycles of EC resulted in a decrease of lymphocyte stimulation whereas 6 cycles of D had no effect. The plasma levels of most immune mediators decreased significantly after six cycles EC when compared to baseline. Under the influence of D, the effect was much weaker. The changes of Eotaxin and OPG during the first 6 cycles of D correlated with the RCB-score in the reverse group. A decrease in Eotaxin (p = 0.0136) or in OPG (p = 0.0487) correlated with good response. The in vitro lymphocyte stimulation assay showed that E and D have similar inhibitory effects on lymphocytes.The annexin V/PI analysis confirmed that E more often leads to apoptotic cell death in SK-BR3 cells than D (E:53% vs. D:14%). SK-BR3 cells formed more often polyploid cells when treated with D. This suggests that D induces a regulated form of necrosis whereas E apoptosis. Conclusions: This study is the first to compare the immunomodulatory effect of E and D in BC patients. E inhibits lymphocyte activation in vitro and in vivo and suppresses many soluble immune mediators. This suggests that it is not suited for combination with immunotherapies. D showed a much weaker effect.