The presence of HPV-DNA was determined in tumor biopsies of cervical-cancer patients and in cervical swabs of non-cancer patients from Tanzania, East Africa, by Southern blot hybridization and/or PCR. HPV types 16 and 18 were detected in 38% and 32%, respectively, of 50 cervical-carcinoma biopsies. A consensus primer PCR capable of detecting a broad spectrum of HPV types revealed the presence of HPV-DNA in 59% of 359 cervical swabs of non-cancer patients. Type-specific PCR showed that types 16 and 18 accounted for 13.2% and 17.5%, respectively, of all HPV infections. Therefore we concluded that HPV 18 is more prevalent in Tanzania than in any other geographical location so far reported. The strongest risk factors for the presence of any HPV-DNA in the 359 female non-cancer patients were young age and HIV infection. The epidemiology of HPV types 16 and 18 was found to differ from that of other HPV types, being associated in univariate analysis with trichomonas vaginalis infection, martial status (single/divorced), age at first intercourse, and young age at menarche. However, young age at menarche accounted for most of the effects of all other, variables in multivariate analysis. Of the non-cancer patients, 12.8% had antibodies against HIV I (no patient being severely symptomatic), and HIV infection was highly correlated with the presence of HPV-DNA, especially types 16 and 18. While HPV-DNA of any type was detectable 1.4-fold more often in HIV-positive patients than in HIV-negative patients, evidence of an infection with HPV types 16 or 18 was found 2.2-fold more often in the HIV-positive patients. The HIV-positive women did not show an increased rate of cervical cytological abnormalities as assessed by PAP staining of a single cervical smear, the overall rate of abnormalities being 2.8%. Furthermore, the age-adjusted prevalence of HIV antibodies was found to be considerably lower in 270 cervical-carcinoma patients (3% HIV-positive) in comparison with non-cancer patients. Thus there was no association observable between the prevalence of HIV infections and the frequency of cervical cytological abnormalities or cervical cancer in the setting of this cross-sectional study.