The authors report on 283 primary, non-metastatic, breast cancer cases consecutively referred after surgery and followed-up from a minimum of 10 months to a maximum of 3.5 years. All cases were studied according to the presence of estrogen receptors (ER). ER presence was correlated with age and menstrual status, with ER+ cases more frequent in older patients. No correlation was found between ER and nodal status. Prognosis was evaluated in terms of disease-free survival at 2 years (actuarial method). No correlation between ER and survival was evident for N- cases, whereas a better prognosis was recorded for ER+N+ patients compared to ER-N+, although the difference was not statistically significant. The observed results are compared with recent literature data and agree with other recent reports, which did not confirm the previously undiscussed statement regarding the prognostic role of ER determination. According to these studies and to the present study, the prognostic role of ER determination seems at least questionable and particularly the postoperative adjuvant treatment of ER-N- cases should be reconsidered.