Methods for protein structure prediction are flourishing and becoming widely available to both experimentalists and computational biologists. But, how good are they? What is their range of applicability and how can we know which method is better suited for the task at hand? These are the questions that this chapter tries to address, by describing automatic evaluation methods as well as the world-wide Critical Assessment of Techniques for Protein Structure Prediction (CASP) initiative and focusing on the specific problems of assessing the quality of a protein 3D model.