Objective: To examine the clinical usefulness of the Bethesda classification system of low-grade cervicovaginal cytology as it relates to predicting underlying histology and aiding in triage to colposcopy. Methods: We evaluated 1454 women with abnormal cytologic screening results: 782 with atypical squamous cells of uncertain significance (atypia), 355 with low-grade squamous intraepithelial lesions (SIL) determined by the presence of human papillomavirus (HPV) alone, and 317 with low-grade SIL determined by the presence of cytologic evidence of mild dysplasia (cervical intraepithelial neoplasia (CIN] I) devoid of HPV cytopathologic features. All women underwent colposcopy, directed-punch biopsy or loop electrosurgical excision, and/or endocervical curettage (ECC), as indicated. Results: Women from the low-grade SIL-CIN I referral cytology group were significantly more likely to harbor all grades of biopsy-proven dysplasia than were those from the atypical squamous cells of uncertain significance and lowgrade SIL-HPV groups, which showed no statistical differences. Conclusions: Contrary to the Bethesda system, which combines CIN I and HPV changes because of cytomorphologic similarities, this study suggests that patients with HPV cytologic smears are similar to patients with atypical smears and are less likely to harbor any biopsy-proven CIN lesions than are patients with CIN I cervicovaginal smears. If excluded from colposcopic triage, approximately 5% of patients with atypical cytologic smears from a well-screened population similar to ours will harbor high-grade lesions that may progress during any waiting period. Triage of low-grade cervicovaginal smears based on histopathologic correlation is encouraged.