Residual stresses often cause distortion to occur when machining slender workpieces like those used in aeronautics. This distortion is undesirable and may even cause the workpiece to be scrapped. Identifying these residual stresses during machining is therefore crucial to limit their effect on the final geometry of the part. Objective: The aim of this work is to identify the through-thickness residual stress distribution by processing displacement/strain fields measured on the workpiece during machining. Methods: Digital Image Correlation is employed to measure, between successive milling passes, the displacement and strain fields on the lateral surface of the workpiece. These fields are then processed with the Virtual Fields Method to identify the through-thickness residual stress distribution. Compared to previous studies on this topic, no assumption is made concerning the real through-thickness displacement field. Results: Simulations performed with synthetic data provided by a finite element model show the feasibility of this approach and quantifies its robustness when displacements are affected by measurement noise. Results obtained with this approach in a real case are then compared with their counterparts obtained with another identification technique, which assumes that the workpiece behaves as a beam. Conclusion: This study shows that it is possible to measure the through-thickness distribution of residual stress in slender workpieces during machining by using a classic subset-based DIC software and the Virtual Fields Method to process the displacement/strain maps. This opens the way for future developments aimed for instance at updating in live machining sequences in order to obtain machined parts immune from distortions caused by residual stresses.