To improve productivity and remain competitive, businesses nowadays introduce the process capability index (Cpm) to evaluate the quality of their products in an effort to improve them and cut down on operation costs. This is because Cpm can clearly reflect process loss and yield percentage (yield%) that it is widely used in the industry. When suppliers’ process capability is found to be limited in terms of Cpm, an improvement in product quality is required and the cost of the improvement varies depending on the source of loss. Though Cpm is a very good index for the evaluation of process capability, it is unable to reflect suppliers’ improvement costs. Thus, this paper takes a reduction in the improvement cost into consideration and proposes the process improvement capability index (CPIM). The mathematical programming model is then used to assess the confidence interval of index CPIM to overcome the problem of complicated estimation of index CPIM. With CPIM, manufacturers are able to evaluate suppliers’ ability in process improvement, particularly when the suppliers’ process capability is found to be limited, to effectively reduce suppliers’ improvement costs, to improve the quality of products, to enhance productivity and finally to achieve the goal of sustainable operations.