In the design of components for service at elevated temperatures, fatigue and creep-fatigue due to cyclic loading are some of the most important failure modes, and the assessment of failure life at structural discontinuities is a key issue within the evaluation of the integrity of the components. Since several fatigue and creep-fatigue life evaluation methods have been proposed, this study compares and assesses these different methods by performing fatigue and creep-fatigue tests of perforated plate made of Mod.9Cr-1Mo steel. Multiperforated plate was subjected to mechanical cyclic loading at 550 °C, and crack initiation and propagation at the surfaces of the holes were observed. The stress distribution was varied by changing the hole arrangement and loading level. A series of finite element analyses (FEA) were carried out to predict the number of cycles to failure by the several failure life evaluation methods, and these predictions were then compared with the test results. Several types of evaluation methods that use the elastic FEA were applied, namely, the stress redistribution locus (SRL) method, simple elastic follow-up method, and the methods described in the design and construction code for fast reactors (FRs) published by the Japan Society of Mechanical Engineers (JSME FRs code). In addition to these, evaluation was also carried out using the results of inelastic FEA to compare these elastic FEA-based estimation methods. The comparisons indicate that, for all conditions tested, the SRL method provided a rational prediction of the fatigue and creep-fatigue life when κ = 1.6 was applied, where κ = 1.6 is the recommended reduction factor for this method in general use. A comparison of the SRL method and the results of the inelastic FEA indicated that the applicability of the value of factor κ in the SRL method depends on the elastic region remaining in the cross section including the evaluated point and the spread in the plastically deformed region in the specimen.