Human papillomavirus (HPV) is a necessary cause of cervical cancer and its precursors. Increased expression of high-risk hrHPV viral oncogenes in abnormal cells might increase the expression of p16INK4a. We aimed to determine the role of p16INK4a in detecting hrHPV-transformed epithelial cells in liquid-based cervical cytology, and compared the results with hrHPV DNA testing by realtime polymerase chain reaction (RT-PCR). Fifty-seven cytological samples were tested for p16INK4a immunomarker and hrHPV DNA. Test performance of both tests was determined by comparing sensitivity, specificity and predictive values using available histological follow-up data as gold standard. Of 57 samples, 36 (63.2%) showed immunoreactivity for p16INK4a and 43 (75.4%) were hrHPV-infected. A fairly low concordance rate (k = 0.504) between p16INK4a immunolabelling and hrHPV DNA status was noted. For prediction of cervical intraepithelial neoplasia (CIN) II and worse lesions, p16INK4a had a sensitivity and specificity of 93.5% and 60%; whereas hrHPV DNA testing had a sensitivity and specificity of 100% and 20%. Dual testing by combining p16INK4a and hrHPV showed sensitivity and specificity of 100% and 33.3%. In conclusion, p16INK4a is useful in predicting severity of the cytological abnormalities. Although p16INK4a is more specific but less sensitive than hrHPV in detecting high-grade cervical lesions, a combination of both tests failed to demonstrate significant improvement in diagnostic sensitivity, specificity and predictive value. Larger-scale prospective studies are required to assess further whether this biomarker should be routinely used as primary screening tool independently or in combination with hrHPV testing to improve diagnostic accuracy in cervical cytology.