Epidermal growth factor receptor (EGFR), androgen receptor (AR) and 14-3-3 sigma have been reported to be implicated in breast tumorigenesis. Their correlations, however, remain elusive in this condition. In order to examine the correlation of EGFR, AR and 14-3-3 sigma in breast cancer, and analyze their relationships with molecular subtypes of breast cancer and their impacts on overall survival, we immunohistochemistrically detected EGFR, AR and 14-3-3 sigma expression in 139 cases of breast cancer. We found that EGFR expression was negatively correlated with AR (r=-0.223, P=0.008) and positively with 14-3-3 sigma expression (r=0.181, P=0.033). There were significant differences in EGFR and AR expression between different molecular subtypes (P=0.000 and P=0.000 respectively). Kaplan-Meier cumulative survival analysis showed that none of the three biomarkers had significant impacts on overall survival of breast cancer patients (P=0.315, P=0.709, P=0.789 respectively). Univariate survival analysis revealed that tumor size (P=0.044), lymph node status (P=0.006) and clinical stage (P=0.008) were significantly associated with overall survival. Multivariate analysis demonstrated that lymph node status was the only statistically significant independent prognostic factor for overall survival [P=0.006, exp (B) = 1.511, CI (1.124-2.032)]. In conclusion, EGFR expression is negatively correlated with AR and positively with 14-3-3 sigma expression in breast cancer. Furthermore, there are significant differences in EGFR and AR expression between various molecular subtypes of breast cancer. Lastly, EGFR, AR and 14-3-3 sigma have no significant impacts on overall survival of breast cancer patients.