Though in the 1980s, colposcopically-directed biopsy excluded over 90% of CIN 3 and cancer (CIN 3+), recent reviews found sensitivity of colposcopically-directed biopsy for CIN 3+ of 50-65%. Studies from China showed that the sensitivity of colposcopically-directed biopsy for CIN 3+ is higher for large CIN 3+ than for small CIN 3+ and higher for associated high-grade cervical cytology than for low-grade cervical cytology. Colposcopically-directed biopsy excluded over 90% of CIN 3+ in the 1980s because colposcopy clinics in the 1980s evaluated women with high-grade cytology that had large CIN 3+; it no longer excludes CIN 3+ well because current colposcopy clinics evaluate women with low-grade cytology that have small CIN 3+. When colposcopically-directed biopsy is used to exclude CIN 3+ our understanding of the natural history of CIN is skewed, errors occur in defining appropriate screening practice, and inaccurate diagnosis results in incorrect treatment. The impression that CIN is more common on the anterior lip of the cervix is an artifact introduced by the inaccuracy of colposcopy. An unjustified enthusiasm for screening with acetic acid aided visual inspection (VIA) occurred when the sensitivity of VIA for CIN 3+ was inflated by screening studies using colposcopically-directed biopsy as the gold-standard for CIN 3+. To limit the harm of inaccurate diagnosis associated with colposcopically-directed biopsy, at colposcopy we advise random biopsies at the squamocolumnar junction in cervical quadrants without visible lesions and, unless the woman is pregnant, endocervical curettage (ECC). As the diagnosis of CIN 3+ solely by ECC is uncommon in women under age 25, the ECC may be omitted in women under age 25 years. If multiple cervical biopsies are performed, to limit discomfort, a bronchoscopy biopsy instrument which obtains 2-mm biopsies should be used.