Researchers and mediators have long been concerned about coercion, intimidation, and safety threats that could occur in mediation for cases where previous violence between the parties has occurred. Most of the research focuses on screening tools to identify parties at risk. When parties screen positive for intimate partner violence (IPV), some proceed to mediation and some do not, depending on the policies of individual mediation programs. But this misses a step: Some cases may benefit from mediation while others won't, but how can we predict whether mediation will be useful and safe in specific instances? This study uses survey data obtained from parties in domestic relations mediations to examine issues of safety, satisfaction, and settlement in the presence of various IPV behaviors.