Abstract Objectives: Magnetoencephalography (MEG) signals are, on theoretical grounds, thought to be relatively undistorted by the skull in contrast to electroencephalographic (EEG) signals. This assumption was experimentally tested in an animal preparation with a brain similar to the human brain in many respects. Methods: Possible skull effects on MEG were evaluated directly using an in vivo porcine preparation, by measuring the somatic evoked magnetic field (SEF) above the skull with and without the skull under, otherwise, the same condition. Results: The SEF was virtually undistorted by the skull with no obvious visible change in its waveform and amplitude under these two conditions. However, there was some small, but significant attenuation when the skull was removed, the distortion being greater for deeper sources. Conclusion: Our results are consistent with a theoretical expectation that the skull should be virtually `transparent' to the magnetic fields for shallow sources, but less so for fields generated by deeper sources.