Objective This study was initially designed to evaluate the effect of celecoxib on the regimen of 5 fluorouracil, epirubicin, and cyclophosphamide (FEC) combination, followed by docetaxel (T) in neoadjuvant setting. An unplanned preliminary review on safety was conducted after a halt of the study due to the conced potential cardiovascular risk of using COX 2 inhibitors.Methods We studied 23 consecutive cases of operable breast cancer having received four cycles of FEC(500 mg/m2, 100 mg/m2, 500 mg/m2) followed by four cycles of T(100 mg/m2) with concurrent celecoxib (400 mg twice daily) (group A) or same chemotherapy regimen but without concurrent celecoxib (group B). These combined chemotherapies were administered every 3 weeks. The Chi square test or Fishers exact test were used to assess the difference in incidence of limiting hematological toxicites between groups. Results 23 patients (group A: n=12; group B, n=11) received a total of 183 out of 184 planned treatment cycles; one (4%, 1/23) of them omitted the fourth cycle of FEC owing to repeated incidences of febrile neutropenia. Received dose intensity (RDI) for FEC in group A (90%±11%) was higher than that in group B (80%±8%) while RDI for T was similar between group A (93%±8%) and group B (96%±9%). Of the first 91 treatment cycles of FEC, limiting hematological toxicity, severe neutropenia including febrile neutropenia, was significantly different between group A and B [(10.4%, 5/48) vs.( 32.6%, 14/43), P=0.009]. Other toxicities commonly observed in chemotherapy receiving patients were manageable. Conclusions Neoadjuvant use of FEC followed by T with concurrent celecoxib appeared to be safe for treatment of operable invasive breast cancer. The observed lower incidence of chemotherapy induced neutropenia is possibly contributed by the administration of COX inhibitor. We believe that further investigation might provide more evidence on the use of COX 2 inhibitors in breast cancer.