Control valves are extensively used in the process industry and valve stiction limits the performance of control loops. This study aims at the quantification of valve stiction that is valuable for improving the control performance by a subsequent step of valve stiction compensation or valve maintenance. A newly proposed two-layer binary tree data-driven model is adopted to capture the valve stiction; the Wiener model is used to describe the process dynamic nonlinearity. That is, a novel Stiction-Wiener model is formulated and a new identification algorithm is proposed to cope with the stiction parameter estimation in the closed-loop environment. It is verified through a numerical case study that the proposed quantification algorithm can provide the consistent estimates of valve stiction parameters.