MUC1 glycoprotein that is overexpressed in aberrant forms in epithelial cancers has been used for diagnosis, staging and therapy. As normal prostate and prostate cancer tissues express MUC1, it represents a potential target, but MUC1 epitopes specific to prostate cancer have not been well characterized. In order to assess MUC1 epitopes in prostate cancer, and their correlation with Gleason grades, binding of 7 well-characterized anti-MUC1 monoclonal antibodies (MAbs) (BrE-3, SM3, BC2, EMA, B27.29, HMFG-1 and NCL MUC1 core), were studied on a prostate tissue microarray. This microarray contained 197 prostate tissue cores representing: i) normal/benign prostate; ii) prostatic intraepithelial neoplasia and Gleason grades 1 and 2; and iii) Gleason grades 3-5. These MAbs bind the MUC1 extracellular domain, but have variable sensitivity to MUC1 glycosylation. To further characterize the effect of glycosylation on their binding, MAb reactivities with unglycosylated MUC1 core peptide and breast and prostate cancer cell lysates were compared. These studies demonstrated strong binding of BrE-3, BC2 and EMA to the peptide core and recognition by BrE-3, SM3, BC2 and EMA of hypoglycosylated MUC1. The results for the microarray indicated that higher Gleason grades were associated with markedly increased cellular staining by MAbs that preferentially recognize less glycosylated MUC1 (BrE-3, p<0.001; SM3, p<0.004; EMA, p=0.009; and BC2, p<0.001). Staining by MAbs that bind preferentially to hyperglycosylated MUC1 (B27.29, p=0.33; HMFG-1, p=0.89; and NCL MUC1 core, p=0.96) did not correlate with Gleason grade. These results demonstrated that hypoglycosylated MUC1 expression increased with Gleason grade, thus supporting the targeting of hypoglycosylated MUC1 epitopes in prostate cancer for more specific imaging and therapy applications.