A novel approach is proposed in this study, which is to evolve visual inspection programs for automatic defect detection on populated printed circuit boards. This GP-based method does not require knowledge of the layout design of a board, nor relevant domain knowledge such as lighting conditions and visual characteristics of the components. Furthermore, conventional image operators are not required to perform the detection. The experiments show that these evolved GP programs can identify all the faults while some suspicious areas are also highlighted. By this GP approach, manual inspection effort can be dramatically reduced. In addition, an evolved GP detection program can readily work on different types of boards without re-training