Generation of macro operators has long been known to be an effective speed-up learning technique for classical planners. It is to be expected that this benefit will carry over to decision-theoretic planners as well. But in decision-theoretic planning additional benefits can be gained from the use of macro operators due to the necessity of comparing alternative plans and the extremely high cost of plan projection. Projecting a probabilistic plan yields an exponential number of outcomes as a function of plan length. Trade-offs between the accuracy of projection and its cost are therefore necessary. One way to do this is to reduce the set of actions and/or the set of branches in each action by creating abstract actions. In a previous paper , we presented methods for abstracting alternative actions to reduce the set of actions, and abstracting groups of branches within an action. In this paper we show how abstract macro operators can be generated which compactly represent a sequence of actions and thus reduce the cost of projection.