Patient management problems (PMPs) have been used for many years in an attempt to design tests capable of measuring clinical problem-solving ability. Problems have arisen particularly related to criterion validity, though some studies have suggested that content and construct validity are satisfactory. One of the major problems has been that a common definition of problem-solving ability has not been developed yet. In addition, various formats have been used with considerable variety in construction, possibly influencing outcomes, and scoring procedures have not been standardized. A serious charge has been that PMPs are content specific and therefore are unreliable and not valid for measuring any general characteristic such as problem-solving ability (assuming this exists).