© 2014 by The American College of Obstetricians and Gynecologists. OBJECTIVE: A post hoc analysis to determine the diagnostic yield of random biopsy in detecting high-grade cervical disease in women with negative colposcopy.METHODS: The ATHENA (Addressing the Need for Advanced HPV Diagnostics) trial screened more than 47,000 women with cytology and high-risk human papillomavirus (HPV) DNA genotyping. Colposcopy was performed in all women with abnormal cytology or positive HPV results. A single random biopsy was taken at the squamocolumnar junction if colposcopy was adequate and no lesions were identified.RESULTS: The random biopsy diagnosed 20.9% (81/388, 95% confidence interval [CI] 16.9-25.3%) and 18.9% (45/ 238, 95% CI 14.1-24.5%) of the total cervical intraepithelial neoplasia (CIN) grade 2 or worse and grade 3 or worse, respectively. This additional disease was detected in both HPV 16 or 18+ and for 12 other high-risk HPV+ women. For HPV 16 or 18, the absolute risk for detection of CIN 2 or worse on random biopsy in the overall population was 13.1% (40/305, 95% CI 9.8-17.4%) and 8.2% (25/305, 95% CI 5.6-11.8%) for CIN 3 or worse. By contrast, the absolute risk for 12 other high-risk HPV+ women was 3.5% (29/820, 95% CI 2.5-5%) and 1.7% (14/820, 95% CI 1.0-2.8%) for CIN 2 or worse and CIN 3 or worse, respectively.CONCLUSION: A single random biopsy increased the detection of high-grade disease when no lesions were visualized at colposcopy. The absolute risks of disease associated with the random biopsy were highest for women positive for genotype 16 or 18. Our study supports performing a random biopsy in women undergoing colposcopy without visible lesions, particularly in those positive for HPV 16 or 18.