A study was conducted to determine the optimal inspection parameters such as range and time for finding defects in carbon/epoxy composite panels using IR thermography imaging. The present paper presents an innovative method for automatically selecting these parameters for evaluating composites based on a series of finite-element models. Such finite-element models of composite panels with flaws at different depth locations were constructed a priori and analysed to estimate the optimal operating parameters. The optimal inspection range and time were identified in the contour plots obtained from the appropriate finite-element analysis results. A graphite–epoxy composite panel with phantom defects at various depth locations was built, and experiments were performed using a thermographic system to verify and validate the proposed method.