In this paper, we study the fulfillment strategies of an omni-channel retailer that would like to leverage its established offline retail channel infrastructure to help its online sales. We consider a single newsvendor-type product that is sold in both online and offline channels to non-overlapping markets with independent Poisson demand. The offline store can fulfill online demand at an additional handling and fulfillment cost, k, but not vice versa. The retailer makes decisions at three different levels: 1) at the strategic level the retailer must establish a fulfillment structure between the two channels in terms of where to stock inventory in the two channels, 2) at the tactical level, the retailer decides how much inventory to have for each channel before the season starts, 3) at the operational level throughout the season, as demand unfolds and inventory depletes, the retailer makes rationing decision about whether to use offline inventory to fill online order at any moment. We build separate and integrated models to study these decisions, and find that the optimal rationing decision has a threshold-based structure that depends critically on k and the mix of demand between the two channels. Two simple rationing heuristics are proposed and shown to be effective. Furthermore, integrating the rationing policy into higher-level decisions, we show that it can have significant impact on the retailer’s stocking and fulfillment structure decisions. As a result, we propose an integrated policy, where the retailer builds separate inventory stocks for each channel but can use the offline inventory to back up online sales, subject to a rationing heuristic, is proved to be simple, effective, and robust. We discuss the various practical implications of our findings.