The newly introduced Bethesda System proposes to replace the terms dysplasia, carcinoma in situ (CIS), or cervical intraepithelial neoplasia (CIN) with two terms, low-grade squamous intraepithelial lesion (SID or high-grade SIL. We tested the biologic relevance of the Bethesda System (ie, its ability to establish a close correlation between biologic behavior and the different grades of lesions) in a series of 528 women with genital human papillomavirus (HPV)- associated precancerous lesions prospectively followed for 10 years. The cervical biopsies were reclassified as either low-grade or high-grade SIL, and the data obtained by colposcopy, Papanicolaou smear, and HPV typing, as well as the biological behavior of the lesions, were analyzed in these two groups. Altogether, 77.4% (376 of 486) of the lesions were classified as low-grade SIL lesions; the rest (22.6%) belonged to the high-grade SIL category. In the low-grade SIL category, 46.8% of the women were 24 years old or younger, as compared with 37.3% in the high-grade SIL group. The colposcopic appearance was normal significantly more frequently in the low-grade SIL lesions (22.1%) than in the high-grade category (8.5%) (P < 001). A single Fapanicolaou smear was inadequate to distinguish between low-grade and high-grade SIL, as evidenced by almost identical distributions of Papanicolaou smear class I and II cells in both categories. Noteworthy was the discovery of normal Papanicolaou smears in 8.2% of the high-grade SIL lesions. Four of the HPV types analyzed (6,11,18, 33) were similarly distributed in both categories, whereas HPVs 16 and 31 were significantly more frequent in high-grade as compared with low-grade SIL (P <.001). The spontaneous regression rate was significantly higher in low-grade SIL (63.6%) than in high-grade SIL (38.2%). Progression was significantly higher (39.1%) in the high-grade SIL category (P<.0001). Despite its definitely positive goals, the drawbacks of the Bethesda System still outnumber its advantages. Most important, simplification of the classification into two grades instead of three in the CIN terminology inevitably leads to significant loss of diagnostically and prognostically valuable information. From the patient's point of view, it is essential that clinical terminology does not result in overtreatment, which seems to be a definite risk using only the two categories of the Bethesda System. The present results firmly advocate preservation of the current descriptive terminology, which has proven its effectiveness for many years.