In robust optimization, we try to find solutions that are good enough with respect to all or most possible scenarios. One way of finding such solutions is via minimizing the maximum regret over all possible scenarios. In this paper, we discuss general results concerning problems with max-regret minimization structure.