Introduction Corrosion of carbon steel in soil is considered to be similar to that in a neutral solution. That is, soil corrosion is considered to be promoted by water and dissolved oxygen in soil water. In the outdoor environment, soil water content changes due to rainfall. Therefore, in order to understand the corrosion mechanism, it is important to evaluate soil corrosion with a system that reproduces the dry and wet cycle. We measured the time course of the corrosion rate in terms of soil corrosiveness in a single wet and dry cycle. So far, we have focused on the soil particle size as the dominant environmental factor, which is considered to affect the diffusion of water and oxygen in the soil, and have clarified the corrosion tendency. In this study, the relationship between the particle size distribution and corrosion behavior was evaluated using the AC impedance method. Experimental The soil was prepared from three levels of blown lowland soil (Soil A, B and C) with different particle size distribution. The particle size distribution was measured by the laser diffraction/scattering method. The soil was placed in a glass container, and the electrodes were buried in the soil. At the beginning of the experiment, enough water was supplied so that the soil was immersed. The soil water content decreased with time after the water supply had been stopped. The impedance measurements were carried out at constant time interval. The temperature in the chamber was held at 25ºC. Results and discussion Figure 1 shows Nyquist plots obtained from the impedance measurement of the carbon steel in soil A at different time. The equivalent circuit is also shown, where R s is the solution resistance, R ct the charge transfer resistance, CPE the constant phase element and Z W the Warburg impedance. Capacitive loops appeared on the Nyquist plots, and the diameter of the loops varied with the measurement time. R ct was calculated from the radius of the capacitive loops. Figure 2 shows the particle size distributions measured at three levels of soil. The maximum frequency appeared around 10 μm in all soils. Soil A contained more particles larger than 100 μm, although the peak at 10 μm was the lowest. On the other hand, Soil C showed the highest peak at 10 μm, but contained few particles larger than 300 μm. Soil B showed properties intermediate between soil A and C. Figure 3 shows the time course of 1/R ct for each soil. Soil A and B showed a 1/R ct maximum around 200 hours, and the maximum of soil A was greater than that of soil B. On the other hand, soil C did not show a clear maximum value of 1/R ct and remained constant. The difference in the corrosion behavior in each soil is considered to be due to the filling rate of the soil particle. It is known that the larger the particle size distribution is, the larger the filling rate of the particles and the smaller the gap diameter between the particles become. In soil with a wide particle size distribution, water is strongly trapped by capillary force in the narrow gaps between particles. Thus, as the soil dried, a thin water film was maintained on the steel surface, and oxygen was efficiently supplied. Therefore, it is considered that the maximum value of 1/R ct was the largest in soil A. Soil C had a high porosity, and the oxygen was efficiently supplied to the steel surface during the drying process, but at the same time, the wetted area decreased. Therefore, 1/R ct did not show a maximum value and remained constant. Conclusion The corrosion rate of steel in soil with different particle size distributions was evaluated by the AC impedance method. It was found that the corrosion behavior differed depending on the particle size distribution. The corrosion behavior is considered to be due to the diffusion of water and oxygen through gaps between particle. Figure 1