Reliable assessment of the epidemiology of genital HPV infections is hamphered by a number of technical problems. Because of the lack of tissue-culture systems, methods based on morphological approaches (colposcopy, cytology and histopathology) play a central role in HPV diagnosis. Even DNA-hybridization techniques and the recently introduced DNA amplification with PCR are extremely difficult to standardize, and are thus subject to major interlaboratory variation. Further confusion in the field is created by the complex biological behaviour of HPV infections. As established by the long-term prospective follow-up study of over 500 women which has been running in Kuopio since 1981, clinical progression and regression are significantly related to the grade of the lesion at the time of diagnosis (p less than 0.00001, and p = 0.0005, respectively), as well as to the type of HPV (p = 0.0012). Most importantly, however, genital HPV infections seem to run an extremely fluctuating course, passage from manifest to subclinical or latent infection being frequently encountered in individual patients when examined at 6-month intervals over prolonged periods. This explains the significantly divergent prevalence figures reported in different series (ranging from 2% to 80%), which are completely dependent on the technique used to analyse the presence of HPV, i.e. whether a) PAP smear, b) biopsy, c) DNA hybridization, or d) PCR amplification. The first two are capable of disclosing only manifest (clinical) infections, the latter two also the latent ones. In an unselected population of 22-year-old Finnish females, the prevalence of clinical HPV infections was about 3 per cent, and the adjusted annual incidence was 8.0 per cent. According to estimates of the life-time risk, up to 79% of Finnish females will contract at least one HPV infection between the ages 20 to 79 years. When related to the long-term trends in invasive cervical cancer in Finland, it is evident that this 79% life-time risk of becoming HPV-infected or even the observed 15% clinical progression rate for HPV infections in the prospective follow-up study by no means signifies an identical risk of developing cervical cancer (i.e. 0.79 x 0.15 = 11%). It seems likely that in countries where mass-screening programmes exist (and precancer lesions are traced), the high prevalence of HPV infections is not necessarily reflected as an increased prevalence of invasive cervical carcinomas. The distinction of lesions at risk for malignant transformation from those regressing spontaneously will have major implications in therapeutic considerations of genital HPV infections.