Due to ever slimmer and higher load-bearing structures the requirements on building materials are increasing. On the part of concrete, the development is therefore moving towards high-strength and ultra-high-strength concretes. In addition to the increasing static stress, the importance of fatigue strength is also increasing due to increasingly sophisticated constructions. Therefore, the focus in materials research is currently on resistance to cyclic stresses, especially in the area of high-performance concretes. Various reasearchers has been detected a heating of test specimens at higher load-speed during pressure swell tests to generate Wöhler lines. For this reason, this study is focused on the heating in relation to ultra-high-strength concrete. Using a comprehensive parameter study, an overview of the significant influencing variables on the heating process could be given. On the one hand, an internal friction potential which increases with decreasing maximum grain size and due to growing damage, could be indetified as an important causes of temperature generation. On the other hand, the applied energy per load cycle is decisive. Unlike the fatigue strength of concrete, which mainly depends on the maximum stress, the heating per load cycle is dependent on the amplitude. The load frequency only influences the measurable heating by changing the time period available per load change for temperature release. But the heating per load cycle is independent of the load frequency. A negative influence of the specimen heating could be observed in the significant reduction of the number of cycles to failure compared to tests in which there is no significant increase in temperature. Based on previous studies on high-strength concretes, various authors propose an adaptation of the test procedure to minimise the temperature development in the specimen. The present work proposes a method in which heating is allowed in favour of time-efficient testing and the maximum temperature is taken into account in the results. The static compressive strength, which is temperature-dependent, could be identified as a main cause of premature failure in the case of strong heating. If the temperature increases, the compressive strength is reduced simultaneously. This leads to a change in the related stress cycle in force-controlled pressure swell tests with constant load cycle. The increasing related maximum stresslevel causes finally a premature fatigue failure. All tests that fail before the calculated expected value heat up until failure. This leads to a permanently changing stress amplitude over the duration of the test. In the evaluation, a changeable load cycle cannot be used for the classification in Wöhler diagrams. Due to the fact that the use of the load input values leads to an underestimation of the fatigue strength, an equivalent constant stress cycle must be determined, which takes into account the strength change of the concrete. Based on the pressure swell tests carried out and the temperature-dependent compressive strength, an analytical method was developed. Using the initial load amplitude as well as the measured maximum temperature, an adjusted maximum stress level can be calculated. The achieved number of cycles to failure can be entered in a Wöhler diagram with the calculated maximum stress level. This method is finally verified for the ultra-high strength concrete investigated in further detail for a wide range of load configurations and additionally for test results of a high-strength concrete.