Consumers need to make prediction on the quality of unused Web services before selecting. Usually, this prediction is made based on other consumers’ experiences. Being aware of different QoS (quality of service) experiences of consumers, this paper proposes a QoS prediction approach. This approach makes similarity mining among consumers and QoS data, and then predicts the QoS of the unused Web services from other consumers’ experiences. Experimental results show that with this approach the preciseness of QoS prediction for Web services can be improved significantly.