It has been previously reported that the magnetic resonance sounding (MRS) method does not produce reliable data in areas where magnetic rocks perturb the geomagnetic field. The applicability of the MRS can be extended by using the spin echo (SE) measuring technique instead of the commonly used measuring scheme based on recordings of the free induction decay (FID) signal. Modifications to the MRS method are presented for measuring and interpreting SE signals. Field results obtained in Cyprus (1999), Canada (2008), and India (2008) reveal that in test sites MRS measurements in the SE mode make it possible to apply the MRS method where the subsurface is composed of sand and gravel that contain magnetite or basalt and in aquifers composed of nonmagnetic sand overlying a magnetic basement. Con-sidering the widespread occurrence of magnetic rocks, this development increases the area where MRS can be applied. However, experience shows that it is more time consuming to measure the SE and more complicated to interpret the field data than it is to work with FID measurements. Numerical results show that the MRS method in the SE mode is less efficient than the FID technique because of the smaller amplitude and wider band of the SE signal. Due to instrumental limitations and unknown distribution of the magnetic fields within the investigated volume, accuracy of the presented MRS-SE approach is site dependent. In a general case, MRS-SE in its current implementation is not able to provide robust estimates of the initial amplitude, which renders MRS estimate of the water content qualitative. For accurate estimate of the water content, more sophisticated approaches need to be developed.