Studies predominantly performed on cystectomy specimens have shown that antibodies against smoothelin can distinguish between muscularis mucosae (MM) (negative or weak stain) and muscularis propria (MP) (strong stain). However, studies on diagnostically difficult (transurethral resection) specimens have not been performed. We studied 34 transurethral resection cases where outside pathologists questioned the presence of MP invasion. Upon expert review of the H&E slides, there was no MP invasion in 18 cases. Smoothelin in MM was negative in 8/18 (44%), weakly positive (1+) in 5/18 (28%), moderately positive (2+) in 4/18 (22%), and moderately/strongly (2-3+) positive in 1/18 (6%). Smoothelin in uninvolved MP present in 8 cases was: 2+ in 2/8 (25%) and 3+in 6/8 (75%). Smoothelin expression in MM was weaker than in MP in 7/8 (88%) cases where both were present. Of 16 tumors with MP invasion, smoothelin in involved MP was: 1+ in 1/16 (6%), 2+ in 3/16 (19%), 2 to 3+ in 9/16 (56%), and 3+ in 3/16 (19%). Smoothelin expression in concurrent uninvolved MP was similar. Our data confirm the relatively distinct staining pattern of smoothelin between MM and MP. However, due to the overlap of intensity between MM and MP, caution should be maintained while using smoothelin immunohistochemistry as a diagnostic tool for MP invasion.