A new classification based on gene expression profiling or immunohistochemical (IHC) characteristics may replace current histopathological classifications and predict better clinical outcomes. We used IHC markers to classify incident cases ascertained by the Palermo Breast Cancer Registry (2002-2004) into four subtypes: luminal-A (ER+ or PgR+ and HER2/neu-); luminal-B (ER+ or PgR+, HER2/neu+); basal-like (ER-, PgR-, HER2/neu-); and HER2+/ER- (HER2/neu+, ER-, PgR-). We evaluated HER2/neu, ER and PgR in 1300/1985 (65%) cases. The most common IHC-subtype was luminal-A (68%), whereas luminal-B, basal-like, and HER2+/ER- accounted for 14%, 13%, and 5%, respectively. IHC-subtypes were not associated with tumor size, geographic location within the province, or menopause, but differed by NPI (P < 0.0001), grading (P < 0.0001), lymph-node involvement (P= 0.04), metastases (P= 0.04), and TNM stage (P= 0.04). Endocrine therapy was administered to 81% of 519 postmenopausal, luminal-A, and luminal-B cases and to 32% of 114 postmenopausal, basal-like, and HER2+/ER- cases.