The increasing retail industry in Indonesia makes competitiveness among retail businesses even higher. Therefore, a good marketing strategy is important to do. Retail Z is a retailer that sells handphone (HP) and electronic accessories, which want to increase sales with limited capital. One of strategy that can be done is product bundling. However, a product bundling strategy that is incompatible with customer needs will be in vain. Therefore, a technique is needed to determine which product will be bundled. By using one of the data mining methods namely Association Rule-Market Basket Analysis it can be determined which products are often purchased simultaneously by taking into account the parameters of support and confidence. Results obtained with the help of RapidMiner software, department 5 and department 2 are departments whose products are often purchased simultaneously. This is indicated by Lift Ratio 1.029, Confidence 0.852 and Support 0.397. So that product bundling can be done by giving discounts to buyers who make product purchases in department 5 and in department 2 simultaneously.