Biomagnetic signals such as MEG (magnetoencephalography) and MCG (magnetocardiography) are very weak compared with those of environmental magnetic fields due to trains and motor cars. In Japan, the fluctuation and frequency dependence of the environmental magnetic noises at extremely low frequencies in urban areas are considerably different between the daytime when direct-current (DC) electric railcars are running and the nighttime when there is no traffic. In order to reduce those noises, magnetically shielded rooms (MSR) should be installed. The shielding effects (SE) of an MSR should be determined by the S/N ratio of the biomagnetic signals to the environmental magnetic noises. Therefore, it is important to understand and predict the characteristics and the amplitude of the fluctuation of the magnetic fields (AFM) at extremely low frequencies resulting from the DC electric railcars. In this paper, firstly, AFM’s within and outside of two kinds of MSR at an individual location are compared to each other in order to confirm that the AFM in a MSR depends not only on the environmental magnetic fields but also on the SE of the MSR. Secondly, the fluctuations of the magnetic fields were measured at three locations near railroads of DC electric railcars. The fluctuations of the magnetic noise at these three measurement points are mainly due to the DC electric railcars. The difference of the amplitude of the fluctuation of magnetic noise depends on the traffic density and the leakage resistance between the rail and the ground. Thirdly, the relationship between the AFM and the distance from the railroad was investigated. The decay factor can be obtained from the relationship between the AFM and the distance from the railroad. These decay factors also depend on the leakage resistance. The decay factors measured are in good agreement with these calculated. By comparing the AFM’s measured as a function of the distance from the railroad with those analyzed as a function of the leakage electric current to the ground, a method of predicting AFM’s due to the DC electric railcars was proposed.