In the February 2008 issue of this journal Allan et al. reported a high prevalence of human papillomavirus (HPV) among South African women (1). We conducted a small pilot study of human immunodeficiency virus (HIV)-infected and non-HIV-infected South African women and found a similarly high HPV prevalence. The real-world impact of the new HPV vaccines will depend in large part upon the distribution of high-risk HPV (HR-HPV) genotypes in a given population. The impact of HPV vaccination will diminish as the proportion of cervical disease that is caused by nonvaccine HR-HPV increases. Regionally Africa has the lowest proportion of cervical cancer cases attributable to the HPV vaccine types 16 and 18 (HPV-16 and -18) (6). Among cytologically normal women Africans have been found to be more likely than non-Africans to be infected with HR-HPV other than HPV-16 (2). Likewise HIV-infected women harbor a distribution of HRHPV that differs from that of the general population and have a decreased relative prevalence of the vaccine genotypes (3 5 8). Given the high prevalence of cervical cancer among women from sub-Saharan Africa and HIV-infected women these subgroups stand to gain the most from effective HPV vaccination. Detailed knowledge of the distribution of HR-HPV among vulnerable populations will inform vaccination strategies. In June 2008 in a prospective feasibility study we collected cervical specimens from 10 HIV-infected women (9 on antiretroviral therapy; mean age 28.1 years; most recent CD4 count average 275 cells/mm3 [interquartile range 176]) and 10 non-HIV-infected women (mean age 28.5 years) attending a health facility in Soweto South Africa. Specimens were collected using the Papette device and stored in vials of ThinPrep solution for transport to our laboratory at the University of Cape Town. HPV genotyping was performed using the Roche Linear Array assay. All 20 samples were b-globin positive confirming adequate DNA collection. Overall we found a 65% prevalence of HPV among our study participants. Thirty infections were identified among the 20 specimens 15 of which were with HR-HPV. Among the 10 samples from HIV-infected women 90% were positive for at least one HPV genotype and 80% were positive for at least one HR-HPV genotype. Multiple infections were identified in 70% of the samples with an average of 2.4 infections per specimen. Among the 10 samples from non-HIV-infected women 40% were positive for at least one HPV genotype and 30% were positive for at least one HR-HPV genotype. Multiple infections were less common than among HIV-infected women (20%) with an average of 0.7 HPV infections (two-sided P = 0.017). A significant difference in the prevalence of HPV-53 (a probable HR-HPV [7]) was identified between these two groups (P = 0.09). Interestingly not a single positive test for HPV-16 or -18 (the two genotypes that are thought to account for approximately 70% of cervical cancer worldwide [4 6 9]) was identified among our 30 positives. Allan et al. identified HPV-35 with the same frequency as HPV-16 among women with high-grade squamous intraepithelial lesions (1). Two of our twenty specimens tested positive for HPV-35. Ultimately however the preventive potential of the current vaccines among South African women will depend upon the HR-HPV distribution among those with significant cervical disease. Our new data are limited in their precision due to the small sample size but support the recommendation that larger studies be conducted on cervical disease in this region. (full-text)