Metallic implants like artificial knee joint distort magnetic field of MRI and this distortion causes artifacts which prevent proper diagnosis. Since most of implants are paramagnetic, the magnetic field distortion ($$$\triangle B/B_{0}$$$) does not depend on the magnetic field of MR scanner ($$$B_{0}$$$). Although artifacts of implants are dependent on pulse sequence and the $$$B_{0}$$$, the $$$\triangle B/B_{0}$$$ maps of metallic implants would enable comprehensive comparison of various implants. In this study, we confirmed that the $$$\triangle B/B_{0}$$$ of metallic materials does not change at 0.4 and 3.0 T. We also derived the susceptibility of a metallic rod from the $$$\triangle B/B_{0}$$$ map and compared its value with the known susceptibility of the material.The $$$\triangle B/B_{0}$$$ maps of metallic samples were obtained by using gradient echo imaging with 2 different TEs. 3D imaging was employed to get rid of slice deformation due to magnetic field distortion and the phase at each voxel was calculated from real and imaginary images. Each sample was placed at the center of the phantom composed of agar (1wt.%) in a lower half volume and water in an upper half one. After imaging of the phantom with a sample, the sample was removed and the empty phantom was also imaged to obtain the intrinsic phase due to the magnetic distortion of the empty phantom itself. This intrinsic phase was subtracted from the phase obtained from the phantom with a sample, and this subtracted phase $$$\phi$$$ was used to calculate the $$$\triangle B/B_{0}$$$ using the equation $$$\triangle B = \phi/\left(\gamma TE\right)$$$ where $$$\gamma$$$ is the gyromagnetic ratio. To compare the magnetic field distortion at different filed strength of $$$B_{0}$$$, an artificial knee joint and a metallic rod (diameter = 3.0 mm, length = 20 mm) made of Elgiloy (Co-Cr-Ni alloy) were imaged at 0.4 and 3.0 T MRI with a spatial resolution of 2×2×2 mm³ and imaging parameters shown in Table 1. The difference of 2 TEs at 0.4 T was set to be 7.5 (3.0/0.4) times larger than that at 3.0 T in order to obtain the almost the same $$$\phi$$$. The susceptibility $$$χ$$$ of material of a rod embedded in water whose susceptibility $$$(χ_w)$$$ is -0.9×10^-5 in SI unit can be expressed by the following equation derived from a theoretical magnetic field of a thin rod aligned to $$$B_{0}$$$. $$χ-χ_w=\frac{∆B(r)×4π×L}{B_0×V×[\frac{1}{(r-L⁄2)^2}-\frac{1}{(r+L⁄2)^2}]} (1)$$ The L and V are a length and a volume of the rod respectively, r is a distance from the center of the rod; Eq. 1 is applicable for a position on a line parallel to $$$B_{0}$$$ and crossing the center of the rod. To obtain the value of $$$ χ$$$, $$$\triangle B$$$ of Elgiloy and Ti alloy (Ti-6Al-4V) rods (diameter = 1.6 mm, length = 39 mm) were mapped at 3.0 T.The $$$\triangle B/B_{0}$$$ of an artificial knee (Fig.1) and an Elgiloy rod did not differ for the magnetic field strength as the theoretical $$$\triangle B$$$ is proportional to $$$χB_{0}$$$. The deficit area of the $$$\triangle B/B_{0}$$$ map (Fig. 1) is determined by the image with a larger TE due to phase dispersion in a voxel. This phase dispersion is proportional to $$$TE × B_{0}$$$: 8.8 (22 ms×0.4 T) and 15 (5 ms×3.0 T). Therefore, the deficit area in Fig. 1 is larger for 3.0 T than for 0.4 T. On the other hand, the contours at 3.0 T was smoother than that at 0.4 T (Fig. 1). This is because of the better SNR at 3.0 T. The SNR of 3D imaging is roughly proportional to NEX, $$$B_{0}$$$ and square root of the number of slices: 8.7 and 29 at 0.4 and 3.0 T, respectively. Although the lower T1 value at lower magnetic fields has advantage with regard to the SNR, the SNR at 3.0 T is still larger than at 0.4 T. The deficit area and the SNR of the magnetic distortion mapping counteract along with the magnetic field of MRI. The magnetic field distortion of an Elgiloy rod extended to the wider area than that of a Ti alloy rod (Fig. 2). The susceptibility value of was calculated as 2.6×10^-3 for Elgiloy and 1.9×10^-4 for Ti alloy. In particular, the susceptibility of Ti-6Al-4V has been precisely measured using magnetometers as 1.8 ×10^-4 [2] that almost agrees with the obtained value in this study. The magnetic field distortion mapping is reliable within 6%. We demonstrated that the $$$\triangle B/B_{0}$$$ of a metallic sample does not depend on the magnetic field strength of MRI. The obtained value of the magnetic field distortion is quantitatively reliable.