Experts in forensic psychology must make skilled observations and conclusions, minimally compromised by bias, in order to try and provide reliable and accurate conclusions to the courts. But the field has little data revealing how well forensic psychologists actually perform these tasks, in part because there has been no clear framework for systematic research of their expertise. Therefore, we consider forensic psychological assessments in light of Dror’s (2016) Hierarchy of Expert Performance (HEP). HEP addresses reliability and biasability, both within and between experts, at the levels of observations and conclusions. Applying this framework to forensic psychological assessments reveals a few domains in which there are some meaningful data, particularly addressing reliability between experts in certain types of forensic assessments. But applying HEP reveals more domains in which we lack data addressing fundamental aspects of expert performance, such as reliability at the level of observations, and reliability and biasability within experts. Understanding these strengths and gaps in forensic assessment research should guide testimony of forensic psychologists, policies around forensic assessment, and further research in forensic assessment.