To investigate the significance of immunohistochemical molecular subtyping, we evaluated outcomes of subtypes based on estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2), and Ki-67. Using tissue microarrays, 1006 breast cancer patients between November 1999 and August 2005 were categorized into four subtypes: luminal A (ER+ and/or PR+, HER2-, Ki-67 < 14%), luminal B (ER+ and/or PR+, HER2-, Ki-67 ≥ 14% or ER+ and/or PR+, HER2+), HER2-enriched (ER-, PR-, HER2+), and triple-negative breast cancer (TNBC) (ER-, PR-, HER2-). Demographics, recurrence patterns, and survival were retrospectively analyzed using uni-/multivariate analyses. Luminal A, luminal B, HER2-enriched, and TNBC accounted for 53.1%, 21.7%, 9.0%, and 16.2% of cases, respectively. Luminal A presented well-differentiation and more co-expression of hormone receptors comparing to luminal B. HER2-enriched showed larger size and higher nodal metastasis. TNBC demonstrated younger age at diagnosis, larger size, undifferentiation, higher proliferation, and frequent visceral metastases. The peak of recurrence for luminal A was at 36 months postoperatively, while that for HER2-enriched and TNBC peaked at 12 months. The relapse risk of luminal B was mixed. Luminal A showed the best survival, but no difference was observed between the other three subtypes. When matched by nodal status, however, TNBC showed the worst outcomes in node-positive patients. In multivariate analyses, luminal A remained a positive prognostic significance. Immunohistochemically-defined subtypes showed different features, recurrence patterns, and survival. Therefore, molecular subtypes using four biomarkers could provide clinically useful information of tumor biology and clinical behaviors, and could be used for determining treatment and surveillance strategies.