More than 80,000 prisoners each year are sexually victimized during incarceration, but only about 8% report victimization to correctional authorities. Complicating reporting is the fact that half of the perpetrators are staff members. Given the restrictive and highly regulated prison environment, studies that examine reporting behaviors are difficult to conduct and to date information available relied on those who have reported or hypothetical victimization studies. This study uses an ecological framework and archival data from a class action lawsuit of sexual misconduct to determine predictors of reporting. Relying on a subsample of 179 women, chosen because they have all experienced at least 1 penetration offense, we use bivariate and multivariable mixed effects logistic regression analyses to examine individual, assault, and context-level predictors of reporting on 397 incidents of staff sexual misconduct. The final model revealed that that 6 predictors (age at time of assault, physical injury, multiple incidents, perpetrator with multiple victims, the year the abuse began, and the number of years women have left on their sentence) account for 58% of the variance in reporting. Disclosure to inmate peers and/or family and friends was significant in the bivariate results. These findings indicate the need for stronger and more systematic implementation of Prison Rape Elimination Act guidelines and remedies that create and enforce sanctions, including termination, for staff violating policy and state law. (PsycINFO Database Record