OBJECTIVES--To investigate the relationship between clinical findings and the detection of human papillomavirus (HPV) DNA in a range of anatomical sites in patients with and without anogenital warts. SUBJECTS--Men and women with a clinical diagnosis of anogenital warts, or a current partner with anogenital warts. SETTING--A department of genitourinary medicine in central London. METHODS--The anogenital areas of the patients were thoroughly examined using a colposcope before and after application of acetic acid. Different types of specimens were taken from a variety of anatomical sites. Superficial skin sampling was performed by the application of slides covered with "Superglue" (SG) to clinically normal and abnormal areas of anogenital skin. The presence of human cells in the SG samples was confirmed by detection of the beta-globin gene using the polymerase chain reaction (PCR). HPV DNA was extracted from the specimens and amplified by using consensus primers with the PCR. HPV types 6, 11, 16, 18, 31 and 33 were identified by Southern blotting followed by hybridisation. RESULTS--In women, HPV DNA was detected in 83% of wart biopsies, 29% of cervical biopsies, 36% of cervical scrapes, 25% of urethral loop specimens, 37% of vaginal washes and 33% of rectal swab specimens. In men, HPV DNA was detected in 67% of wart biopsies, 37% of urethral loop specimens and 12% of rectal swab specimens. Of the SG samples containing the beta-globin gene, 49% from women and 50% from men contained HPV DNA. HPV DNA was not detected in buccal scrapes and serum samples from women or men. Of all specimens with detectable HPV DNA, there was evidence of a single HPV type in 41%, multiple types in 48% and undetermined types in 11%. Samples taken from different sites of a patient tended to have HPV types in common. Sexual partners, however, did not consistently have HPV types in common. CONCLUSIONS--HPV DNA was distributed widely in the anogenital area, in warts, acetowhite areas and clinically normal skin. The SG technique was well tolerated by patients and produced results consistent with other findings. Sampling from a single site of the genitalia on one occasion may significantly underestimate the infection rate with HPV. Multifocal infection of the anogenital area with HPV should be taken into consideration when interpreting epidemiological studies and management strategies.