The invention provides an intelligent retail service method and system for accurately recommending related commodities. The method comprises the following steps: S101, acquiring voice data, limb movement data and face data when a customer browses the commodities; s102, feeding back commodity related information to the customer according to the voice data, the body movement data and the face data when the customer browses the commodity, searching the commodity of the category in a commodity database according to the commodity required by the customer, and moving along with the customer in real time; s103, performing voice recommendation on the commodities of the category according to the priority level, and displaying attribute information of the commodities through a display screen; s104, scanning the bar code of the commodity to complete self-service settlement, and adjusting warehouse-in and warehouse-out data of the commodity. The invention can judge the demands of the customer according to the voice data and the body movement data of the customer in combination with the face data so as to accurately recommend the related commodities, can help the customer with difficulty to select the commodities, improves the marketing success rate, saves the time of the customer, and is suitable for the retail field.