In order to establish sophisticated management of aging degradation and to achieve high reliability of components in nuclear power plants, it is required to reveal the mechanism of aging degradation and to quantify its deterioration. In low-cycle fatigue regime, it was shown that the number of cycles to specimen failure (fatigue life) can be estimated by predicting crack growth. The application of crack length for representing fatigue damage will make it possible to measure fatigue damage by inspection. In previous study, crack growth with inhomogeneous rate and fatigue life in a typical condition were predicted with statistical model of micro crack growth. In this study, fatigue test in expanded condition was conducted in air at room temperature in order to define the relationship between damage factor (DF = number of cycles/fatigue life) and crack length. Crack initiation and crack propagation during fatigue test were measured by replica investigation periodically. Statistical model of micro crack growth was applied to the observation result and its applicability for various strain ranges was demonstrated. The relationship between the damage factor and crack length was shown and its use for maintenance was discussed.