During high-speed end mill boring, dynamic scale mismatch originates from the difference in the dynamic response of each axis and, at high feedrates, causes a trajectory ovalization along the axis with the fastest dynamic response. To investigate this error, an experimental approach is used at different feedrates and trajectory radii for a high-speed machine tool with linear motor drives. Results show that dynamic scale mismatch causes out-of-roundness and radius size errors. A simple second order model is used successfully to predict the dynamic scale mismatch and a strategy is proposed and tested to compensate it at the G-code level. An experimental trial reveals the usefulness of the approach.