purpose: to evaluate the loco-regional response to primary chemotherapy in patients with breast cancer at stages ii and iii. methods: a retrospective and analytical clinical study carried out in 97 patients with an average age of 52.2 years old, with breast cancer at stages ii and iii, attended from january 1993 to december 2004, and submitted to 3 to 4 cycles of primary chemotherapy with 5-fluorouracil - 500 mg/m2, epirubicin - 50 mg/m2 and cyclophosphamide - 500 mg/m2 or doxorubicin - 50 mg/m2 e cyclophosphamide - 500 mg/m2, and then to loco-regional surgical conservative or radical surgical treatment. chi-square and fisher？s exact tests were used to study the association among the variables (age, menopausal state, pre-chemotherapy tumoral volume, axillary condition, stage, therapeutic scheme and number of cycles), while pearson？s correlation coefficient was used for the quantitative variables (tumoral volume according to the anatomo-pathological study and the post-chemotherapy clinical tumoral volume. the significance level was 5%. results: there were 56.8% of cases at stage ii and 43.2% at stage iii. approximately 50% of the patients received fec50 and 50% ac. objective clinical response with primary chemotherapy was obtained in 64.9% of the cases. full clinical response occurred in 12.3% of patients, while full pathological response occurred in 10.3% of the cases. conclusions: there was a statistically significant correlation between the number of cycles and the response to primary chemotherapy. patients who received 4 cycles had better response than those who received 3 cycles. there was also a statistically significant concordance between the evaluation through clinical examination of the response to primary chemotherapy and the pathological findings. no statistically significant correlation was observed concerning age, menopausal status, tumoral volume, and pretreatment of axillary damage.