Liquid-based cytology (LBC) enables the use of supplementary methods in the diagnosis and prognosis of cervical lesions. The aim of this study was to analyze the correlation between p16-INK4a immunoexpression in ThinPrep cervical cytologic samples and human papillomavirus (HPV) detection by polymerase chain reaction (PCR) from the same sample. LBC-ThinPrep (Cytyc, USA) cervical cytology samples, prepared and stained by Papanicolaou method, were analyzed using modified Bethesda cytologic classification named "Zagreb 2002". A second ThinPrep slide, prepared from the same sample, was immunostained for p16INK4a using CINtec p16INK4a Cytology Kit (DakoCytomation, Denmark). Increased expression of the high-risk (HR) HPV E6 and E7 oncogenes results in a highly specific increase in p16 protein expression and overexpression of p16INK4a acts as a potential biomarker for cervical cancer progression from premalignant lesions. Brown nuclear and/or cytoplasmic staining of abnormal cells was considered a positive result. Residual material was used for 13 HR HPV-DNA detection by the PCR based AMPLICOR HPV test (Roche Molecular Systems). A total of 120 ThinPrep Pap tests with the following cytologic diagnoses: 17 within normal limits, 17 atypical squamous cell (ASC) (7 ASC of undetermined significance /ASCUS/ and 10 ASC of high-grade squamous intraepithelial lesions cannot be excluded /ASC-H/), 26 low-grade squamous intraepithelial lesions (LSIL) corresponding cervical intraepithelial neoplasia (CIN) 1, 57 high-grade SIL (HSIL) i.e. 24 CIN II and 33 CIN III and 3 squamous cell carcinoma (SCC) were included in the study. All CIN III (n = 33) and SCC (n = 3) specimens expressed p16INK4a immunoreactivity, whereas the HR HPV test was positive in 97% (32/33) of CIN III and 100% (3/3) of SCC specimens. The p16INK4a biomarker was positive in 87.5% (21/24) of CIN II and 69% (18/26) of CIN I, while the HR HPV was positive in 75% (18/24) of CIN II and 50% (13/26) of CIN I. In ASCUS cytology, p16INK4a and HR HPV showed the same rate of positivity (28.5%; 2/7). Expression of p16INK4a was detected in all cytologic (10/10) ASC-H lesions, in contrast to HR HPV detected in only 20% (2/10) of ASC-H cases. These data suggest the p16INK4a evaluation in ThinPrep cervical samples to be significantly associated with HR HPV testing by PCR in the same sample for the diagnosis of HSIL lesions and cervical carcinomas. A prospective study with longer follow up may clarify the predictive values in the management of LSIL and ASC diagnosis.