In mechanical industry products are usually manufactured by assembling a large number of subromponents. The production and stocking of such components are often of much greater significance then the assembly and stocking of final products. The problem is set within a multicriteria framework, and the consequences of the tactical decisions are presented to the decision maker as goal outcomes. The methodology involves embedding relatively simple models within a larger and more comlex model. The large total model is a simulation model which uses the optimal decisions of the inventory model to determine the strategic concequences for each important criterion. The methodology was applied to an inventory system of a manufacturer of spare parts for cars. The results of that application are reported in some detail. Of particular interest is the comparison of two inventory models: the economic order quantity model and a stochastic inventory model. These two models are compared on the basis of their performance in producing desirable strategic consequences.