The aims of the study were to estimate the prevalence of HPV infection in patients treated for high grade lesions of the cervix uteri (HG CIN), and to evaluate the validity of the histological criteria used for detection of HPV infection. The study comprised 203 women treated for HG CIN by laser conization. Forty-three preoperative biopsies and 160 cone specimens were examined for HPV infection using light microscopy (LM), in situ hybridization (ISH), and polymerase chain reaction (PCR). ISH was performed using commercial biotinylated probes for HPV types 6/11, 16/18, and 31/33/35 (Vira-Type In Situ Kits, Digene Diagnostics, Silver Spring, MD). HPV-PCR was performed with the L1 consensus primers Gp5+/6+. The prevalence of HPV detected by LM was 70%; by ISH 48% and by PCR 83%. Using PCR as the gold standard, LM had a sensitivity of 0.71 and specificity of 0.41. The corresponding results for ISH were 0.51 and 0.65, respectively. The positive predictive value for both tests reached over 80%, but the negative predictive value was less than 25%. This study demonstrates that morphology is an unspecific method of identifying HPV infections. LM identification of HPV infections has no clinical implications. Our analyses comparing test performances of LM, ISH and PCR show that PCR is the superior method.