The distinction between an endometrial and an endocervical origin of an adenocarcinoma may be difficult, especially with small biopsy specimens or when tumor is present in both endometrial and cervical specimens. Previous studies have investigated the value of antibodies such as carcinoembryonic antigen, estrogen receptor, and vimentin in making this distinction. We investigated the value of p16 immunohistochemistry for distinguishing between an endometrial and an endocervical origin of an adenocarcinoma. Cases included in the study were endometrial adenocarcinomas of endometrioid type (n=29) and cervical adenocarcinomas of endocervical type (n=23). Cases were scored on a scale from 0 to 5 depending on the percentage of positive tumor cells: 0 (negative or occasional cells positive); 1 (<5% cells positive); 2 (5-20% cells positive); 3 (21-50% cells positive); 4 (51-99% cells positive); 5 (100% cells positive). Twenty-two of 23 (96%) endocervical adenocarcinomas were scored 5; the other was scored 0. The numbers of endometrial adenocarcinomas with scores of 0 to 5, respectively, were 1, 7, 4, 9, 5, and 3. Most primary endocervical adenocarcinomas were characterized by strong, diffuse positivity of 100% of cells with p16. Endometrial adenocarcinomas are usually positive, but positivity is generally focal and commonly involves <50% of cells. However, occasional endometrial adenocarcinomas exhibit 100% positivity. Diffuse, strong positivity with p16 suggests an endocervical rather than an endometrial origin of an adenocarcinoma. When there is morphological doubt this antibody may be of value as part of a panel for ascertaining the origin of an adenocarcinoma. Diffuse, strong positivity with p16 in endocervical adenocarcinomas is likely caused by inactivation of the retinoblastoma protein by the E7 human papillomavirus oncoprotein, which acts as a p16 transcript repressor.