Since 1963 when recording of the human magnetocardiogram (MCG) was first reported, the number of clinical studies has been limited. High-quality tracings can now be easily obtained in special shielded chambers, but problems remain if records are to be made in hospitals without such chambers. Measurement techniques, theory of the MCG, and model studies are discussed. Configuration of the MCG waveform is much the same as that of the electrocardiogram (ECG). Measurements of dc currents of injury in dogs by means of the MCG have been reported, and are potentially of great clinical interest. Effects of magnetic susceptibility changes in the torso associated with blood movement may contribute to variations in the external field. The questions of what new diagnostic information is available in the MCG and what lead systems are most appropriate remain to be answered.