Because of the fact that any meaningful classification should bear a close relationship to the biological behavior of the lesions, the usefulness of all new classifications of cervical precancer lesions can only be established by well controlled prospective follow-up studies. However, several methodological and conceptual problems are encountered in the natural history studies conducted during the past several decades. While reviewing the available prospective follow-up studies on cervical intraepithelial neoplasia (CIN), Ostör (1993) found 3529 cases of CIN 1, of which 57% showed regression, persistence was found in 32%, progression to CIN III in 11%, and progression to invasive cancer in 1% of cases. The corresponding figures for CIN II were 43%, 35%, 22%, and 5%, respectively. The recognition of the association between human papillomavirus (HPV) and CIN has further complicated the assessment of the natural history of cervical precancer lesions. Results from the early prospective follow-up studies are remarkably consistent, however. Progression from HPV-NCIN (i.e., koilocytosis without CIN) to CIN I or greater was reported for 18 (8%) of 232 women followed by Syrjänen et al. for an average of 25 months, for 26 (8%) of 314 women followed by de Brux et al. (1981) for 15 to 18 months, and for 113 (13%) of 846 women followed for up to six years by Mitchell et al. During a 42-month follow-up period, 10% progression rate was found in 1269 women with HPV-CIN I, and in 17% of 762 women with HPV-CIN II by de Brux et al. (1983). The spontaneous regression rates were 53% and 39% in these cohorts, respectively. This is fully consonant with our experience from an almost 14-year follow-up of 530 women in Kuopio, where the spontaneous regression rate seems to increase in parallel with the extent of the follow-up time, currently being 66.7% for HPV-NCIN and 55.7% for HPV-CIN I. The figures for progression are 6.3% and 14.2%, respectively. It is obvious that the probability of a cervical precancer lesion to progress into an invasive disease increases with the severity of the atypia. Another distinct prognostic factor is HPV type, HPV 16 lesions possessing a significantly higher risk for progression than infections by other HPV types. The follow-up data also indicate, however, that even the high grade lesions may spontaneously regress, which should have important implications in therapy. The continuous problem still remains; these natural history observations only apply to a large series of women but are of little help in predicting the disease outcome in individual women.