Purpose: This paper focuses on the PC industry, analyzing a PC supply chain system composed of onelarge retailer and two manufacturers. The retailer informs the suppliers of the total order quantity, namelyQ, based on demand forecast ahead of the selling season. The suppliers manufacture products accordingto the predicted quantity. When the actual demand has been observed, the retailer conducts demandlearning and determines the actual order quantity. Under the assumption that the products of the twosuppliers are one-way substitutable, an integrated decision-making model for dynamic pricing andinventory control is established.Design/methodology/approach: This paper proposes a mathematical model where a large domestichousehold appliance retailer decides the optimal original ordering quantity before the selling season and theoptimal actual ordering quantity, and two manufacturers decide the optimal wholesale price.Findings:By applying this model to a large domestic household appliance retail terminal, the authors canconclude that the model is quite feasible and effective. Meanwhile, the results of simulation analysis showthat when the product prices of two manufacturers both reduce gradually, one manufacturer will often waittill the other manufacturer reduces their price to a crucial inflection point, then their profit will show aqualitative change instead of a real-time profit-price change.Practical implications: This model can be adopted to a supply chain system composed of one largeretailer and two manufacturers, helping manufacturers better make a pricing and inventory controldecision.Originality/value: Previous research focuses on the ordering quantity directly be decided. Limited workhas considered the actual ordering quantity based on demand learning. However, this paper considers boththe optimal original ordering quantity before the selling season and the optimal actual ordering quantityfrom the perspective of the retailer