Psychologists are sometimes asked to provide their expert opinion in court on whether memories of victims, witnesses, or suspects are reliable or not. In this article, we will discuss what expert witnesses can reliably say about memory in the legal arena. We argue that before research on memory can be discussed in legal cases, this research should ideally meet the following three conditions: replicability, generalizability, and practical rele- vance. Using a fictitious false memory case, we offer a guide to how psychologists should critically examine whether a particular segment of memory research is in line with these three conditions. We show that the area of false memory broadly fits these conditions but that for areas such as eyewitness identification and false con- fessions, there is limited discussion on which effect sizes are of interest in legal cases. We propose several rec- ommendations that expert witnesses can use when they evaluate the validity of statements such as working with scenarios (e.g., statements are valid or not). Being transparent about the limits and strengths of memory research will assist triers of fact in their decision-making process.