Tissue biopsy following a pap test diagnosis of high grade squamous intraepithelial lesion (HSIL) sometimes fails to confirm the presence of a corresponding high grade cervical intraepithelial lesion (CIN 2-3), leading to confusion as to how best to manage the patient. It has been shown that these patients are still at higher risk for future detection of CIN 2-3 even if the initial biopsy fails to detect it. It has also been shown that immunohistochemical staining for p16INK4a can be reliably used as a surrogate marker for infection with high risk human papillomavirus in cervical samples, and that it can be used to enhance detection of CIN2-3 in cases where suspicion is high. To evaluate the use of p16INK4a staining in cases of HSIL which were not confirmed on initial biopsy, two pathologists rereviewed Pap and hematoxylin and eosin preparations from all such cases seen within the preceding 3 years. Immunohistochemical study for p16INK4a was performed and graded on representative sections. The results were tabulated and analyzed. Of the identified 596 HSIL Pap cases, 82% had HSIL on initial cervical specimens. Table 1 shows the 56 cases included in the study with graded and stratified p16INK4a results. On review of the p16INK4a slides, only 2 cases could be upgraded to HSIL/CIN2-3 from the original diagnosis. p16INK4a 2-3+ was expressed more frequently in cases initially interpreted on Pap as low-grade cervical lesion as compared with benign (24 of 35 cases). In the younger than 24-yr-old group p16 2-3+ reactivity was more frequent in benign and low-grade cervical lesion/CIN1 groups (benign: 3 of 5 cases, and CIN1: 6 of 8), and p16 negative reactivity was not seen. p16INK4a was graded 0-1+ more frequently in specimens interpreted as benign in the older than 25 yr olds (10 of 16 cases). The study suggests some diagnostic benefit from the use of p16INK4a immunohistochemical study on cervical specimens from women with a HSIL Pap test without HSIL/CIN2-3 on original hematoxylin and eosin review.