Human papillomavirus type 16 (HPV-16) DNA is detected commonly in cervical carcinomas; in this study, we have determined the analytical sensitivities of Hybrid Capture, HPV-consensus PCR, and three HPV-16-specific polymerasechain reactions (PCRs) for the detection of HPV-I6 DNA. Samples investigated included a cervical cancer cell line, cervical scrapes from 20 patients attending colposcopy clinics, and buccal swabs from eight immunosuppressed children. HPV-16 E7 and E5-nested PCRs [Cavuslu et al. (1996): Journal of Virological Methods, in press1 produced positive signals from samples containing fewer than ten HPV-16 genomes per reaction. HPV-consensus PCR [Manos et al. (1989): Cancer Cells 7:209-2141 and HPV-16 PCR using primers of van den Brule et al. [(1990): Journal of Clinical Microbiology 25:2739-27431 were of intermediate sensitivity (i.e., produced positive signals from samples containing 250 and 2,500 HPV-I6 genoms/reaction, respectively) and Hybrid Capture could detect just 50,000 HPV-I6 genomes/ reaction. Highest rates of positivity for cervical samples were detected with HPV-I6 E7 or E5nested PCRs 150% (10 of 20 samples) and 60% (12 of 20 samples) positive, respectively], intermediate rates with HPV-consensus PCR and PCRs using the primers of van den Brule et al. [both 35% (7 of 20 samples)l, and lowest rates of positivity [25% (5 of 20 samples)] with Hybrid Capture. None of eight buccal swab samples from immunosuppressed children were positive by Hybrid Capture, yet three (37.5%) were positive by HPV-16 E5-nested PCR. These data indicate that HPV-16 type-specific PCRs should be used for the investigation of specimens that may contain low amounts of HPV-16 DNA. o 1996Wiley-Liss, Inc.