Abstract: The optimal joint ordering-pricing decisions are studied in an assembly system consisting of n components in a single period setting. Demand for the end-product is random and component supply capacities are uncertain due to unexpected breakdowns, repairs and reworks, etc. The assembler jointly decides components ordering amounts and end-product price before the supply capacities and the demand are realized. Using stochastic nonlinear programming, a multidimensional optimization model in which the object is to maximize the assembler's expected profit is presented. It is showed that the optimal ordering policy is to order equally despite the supply uncertainties, and then the n+1-dimensional model can be reduced tobe of two dimensions (i.e., a unified ordering decision and a pricing decision). One decision is fixed and the corresponding optimal path of another decision is derived, and then the global optimal solution along the path is solved. Numerical experiments are conducted to justify the model and to further explore the effect of random supply and demand. For the random supply, the increase of the mean is related to investments in new production line or promises of overtime work, the decrease of variance is related to enhancements of preventive maintenance for manufacturing equipments or enhancements of training and education for workers to avoid misoperations, etc., the simultaneous increase of mean andvariance is related to supplier outsourcing its work to multiple second-tier suppliers. Based on these relations with practice, logical explanations are came up with and managers are offered with useful suggestions and insights.

Key words: ordering, pricing, assembly system, random supply, random demand