Strain controlled two stage low cycle fatigue tests were carried out, where cycle ratio was changed in five conditions and strain amplitude in three conditions, and the results were discussed including the results of the previous report. Cyclic hysteresis energy in the low cycle fatigue test was analysed theoretically and it was found that the rate of contribution of the hysteresis energy to fatigue fracture changed with strain amplitude and strain amplitude ratio _??_ _{ta ·2}/_??_ _{ta ·1} as parameters. The relation between longitudinal true strain amplitude or plastic true strain amplitude and number of cycles to crack initiation in the second stage was found to be linear for both step-up and step-down regions, with having a knuckle point on the level of the strain amplitude in the first stage. On the other hand, the relation between the strain amplitude and number of cycles to failure was able to be fitted with straight line with almost the same inclinations to that of single stage test. Constants and exponents of Manson-Coffin's equation in the step-up region changed with varying strain amplitude and cycle ratio in the first stage. As a consequence, usage factor on crack initiation bases was expressed in terms of total strain amplitude, total strain amplitude ratio, and cycle ratio. In the step-down region, crack initiation life and failure life in the second stage were linearly related each other, and ratio of crack initiation life to failure life was expressed by taking cycle ratio in the first stage as a parameter.