AbstractIn certain inventory situations it makes sense to coordinate the control of a group or family of items in that a single major ordering (replenishment) cost is incurred if any item in the group is ordered (e.g. where the group of items are all ordered from the same suppli-er). A reasonable control system is a so-called can- order point or (S,c,s) system. In such a system a group replenishment is enacted when any item 1 hits its must- order point, s1. Enough is ordered of item i to raise its inventory position to its order-up-to-level, S1. At the same time any other item J with a low enough inventory level, namely at or below its can-order point ej, is included in the group order so as to raise its Inventory position to Sj. The mathematical analysis of such a system, e.g. the computation of the expected service level of a particular item, is extremely complicated. In this paper we discuss the use of simulation to estimate two common measures of service. Particular attention is devoted to ascertaining the amount of variance reduction achieved through the use of Conditional Monte Carlo.