Abstract A company's manufacturing is often characterized as either make-to-stock (MTS), make-to-order (MTO), or assemble-to-order (ATO). This classification relates to the degree of interaction between the technological core and the market, with MTS involving the least amount of interaction and MTO the highest degree of contact. ATO represents a hybrid manufacturing strategy for which parts and subassemblies are made according to forecasts while the final assembly of the products is delayed until customer orders have been received. It is evident that each manufacturing philosophy has strategic as well as operational implications. This paper focuses on ATO manufacturing and, in particular, on the design and operation of the manufacturing planning and control system. The relative differences between MTO, ATO, and MTS strategies, and the reasons why a company may decide to be an ATO manufacturer, are discussed first. Several problem areas, which must be addressed by a company that chooses this form of manufacturing, are then identified. It is found that the ATO philosophy requires special system design considerations, particularly in the areas of master scheduling, bills of material structuring, order entry/order promising, final assembly scheduling, and buffering against demand uncertainty. Examples of important issues discussed are the selection of appropriate master schedule units and the associated consequences for the structuring of the bills of material; the choice of efficient procedures for reliable order booking using the combined information from forecasts, confirmed orders, and the master schedule; the relationship between the final assembly schedule and the master schedule; and various techniques that can be used to counter the effects of forecast errors related to the demand for customer options. Following the discussion of all the issues raised above and the presentation of some common solutions, actual industry applications taken from several companies are used to illustrate various aspects of ATO manufacturing. These case illustrations complement the other material since many of them were chosen to reflect alternative procedural approaches to some ATO-related problems. Finally, areas for future research into the design and operation of ATO manufacturing systems are suggested. It is particularly noted that few normative models for systems design exist in this area.