The demonstration of human papillomavirus (HPV) in 99.7% of cervical carcinoma surgical specimens from around the world required investigations by multiple alternative polymerase chain reaction (PCR) assays. A similar approach may therefore be necessary to best characterize HPV prevalence and genotype distribution among cervical cytology samples. In an earlier study, 752 of 799 (94.1%) abnormal and 82 of 300 (27.3%) normal cytology specimens tested HPV positive after PCR using GP5+/6+primers. This study has reinvestigated the "HPV negative" abnormal samples (20 atypical squamous cells of undetermined significance, 5 low-grade squamous intraepithelial lesion, 14 atypical squamous cells, cannot exclude HSIL, 6 high-grade squamous intraepithelial lesion) and an age-matched cohort of "HPV negative" normal (negative for an intraepithelial lesion or malignancy) samples by PCR using PGMY09/11, FAP59/64, and LCR-E7 primers. PGMY09/11-GP5+/6+ nested PCR was performed on samples that were HPV negative by PGMY09/11 PCR. After the first 3 assays, HPV was detected in 41 of 45 (91.1%) abnormal and in 10 of 47 (21.3%) normal samples (P<0.0001). Eighteen HPV genotypes were detected and in some samples the genotype that was identified differed between the tests. The nondetection of common HPV genotypes (eg, HPVs 6, 11, 16, and 18) was notable. High-grade histopathology was found for 2 patients with HPV52-positive cytopathology. Combined with our earlier study, HPV (40 different genotypes) is shown in 99.5% of abnormal samples (99.8% inclusive of the nested PCR data). These findings show that HPV genotype and prevalence estimates are dependent on the method(s) of detection and indicate that suboptimal analytical sensitivity for one or more of the less common high-risk HPV genotypes could lead to impaired clinical sensitivity. HPV may be causal in almost every instance of abnormal cervical cytology; however, passenger HPV that is incidental to an abnormality may also have been detected.