The first part of this diploma thesis deals with choosing and establishing a suitable database for testing results (low and high cycle fatigue) considering the requirements of the Chair of Mechanical Engineering. This is an important point, because every year hundreds of material and component part experiments are carried out. Therefore lots of testing data is produced and needs to be stored in an intelligent way in order to provide an easy access later on. The literature study shows that a relational database including a three layer structure (projects, test series, specimen) is the most suitable solution for the given task. A prototype of the intended database was set up, which will be developed further on through an external specialist. In the second part of the thesis several methods for approximately locating the fatigue limit, which is used in planning Wöhler-Tests, were developed and tested. This task demands special attention, because it is the base for choosing the load levels for testing. All methods were applied on eight different materials. A comparison of the results shows that the choice for selecting one method for a given material depends on the class of material and on the desired result (only compression fatigue limit or compression and tension fatigue limit). It is recommended to use either the FKM or Hück formulas, the Locati method or the recharge procedure. This section also includes a full guideline for planning Wöhler-Test by using these methods and shall be a support for the employees of the chair. The third part is concerned about improving the commonly used evaluation methods for Wöhler-Test results at the Chair of Mechanical Engineering. These are the log10 normal distribution or graphic evaluation for the endurance limit and the Arcsin√P or Probit Methods for the fatigue limit. It was shown that the currently applied formula for estimating the possibility of fracture and the standard deviation have improper outcomes. Therefore several new approaches found in publications were presented and discussed.