Ball-end mills are commonly employed for semi-finishing and finishing of complex surfaces, widely utilized in the energy, automotive, aerospace and other industries. During surface machining, cutter workpiece engagement (CWE) area of the ball-end mill continuously varies with surface characteristics and the tool paths. Additionally, the actual position of the tool involved in cutting also changes, leading to uneven wear distribution on the ballend mill, making tool wear prediction and evaluation challenging. In this paper, a discretization calculation method of tool path length for surface machining is proposed based on PowerMill post-processing development. The linear cutting length of tool path for each corresponding tool posture is calculated, and the approximate calculation model of average milling distance of single tooth for cutting edge element is established. The correlation mapping between NC machining program, tool posture and cutting workload of cutting edge element is realized. A tool wear model that considers effective cutting distance is developed, enabling the prediction of wear distribution along the flank of the ball-end mill under the surface machining conditions. Comparative experiments were conducted on tool wear under different feed directions during curved surface machining. The experimental results confirm the effectiveness of the proposed ball-end mill wear prediction method.