Fatigue in welded steel structures, such as bridge structures, crane runway beams, offshore structures, wind energy supporting structures, etc., is a progressive and localized process of damage accumulation in the material due to cyclic stresses. The magnitude of cyclic stresses is often below the material yield strength. Fatigue damages in steel structures most often occur in geometric and material discontinuities such as welded steel details. Welding affects the material properties, which can cause inhomogeneity within the welds, such as notches, pores, voids etc. Weld also represents a sudden change in the geometry of the detail that causes high-stress concentrations. Welding is performed by melting the base and additional material using concentrated heat, which causes high tensile residual stresses after cooling in the heat-affected zone. Geometrical, material discontinuities and high residual stresses significantly reduce the fatigue life of welded steel details. There are several methods to increase the fatigue life of such details. Fatigue damage can be avoided by applying lower-stress concentration details or by positioning welds in the lower-stress areas in the structure. High-quality welds without voids and imperfections and full penetration welds can also increase fatigue life. Despite design guidelines, especially in the case of high-strength steel applications, it is not always possible to achieve sufficient fatigue strength. Therefore, Post Weld Treatment methods have been developed, such as the High-Frequency Mechanical Impact method, which can achieve a significant increase in the fatigue life of welded steel details. Fatigue life improvement is achieved by improving the weld geometry, reducing the stress concentration, introducing compressive residual stresses and increasing the hardness at the weld toe. These are the HFMI improvement parameters. The most common method for fatigue life assessment is the S-N method, which considers the total fatigue life of welded steel details. The fatigue life of steel structures consists of a crack initiation and crack propagation period. While unwelded details show more extended periods of crack initiation, in the welded details, the period of crack propagation has a dominant influence. HFMI treatment extends the crack initiation period of a welded steel detail, so it would be appropriate to consider these two periods separately. This research is based on developing and calibrating deterministic and stochastic calculation models for fatigue life assessment of as-welded and High-Frequency Mechanical Impact treated details. Calculation models are suitable to consider both crack initiation and propagation separately and provide a reliable fatigue life assessment of the considered details. To develop a reliable model, it is necessary to calibrate it with specific laboratory fatigue tests where all HFMI improvement parameters will be measured on the considered specimens. As-welded and HFMI-treated test specimens were also 3D scanned to obtain a precise geometry of welded steel details. Quality control assurance for HFMI treatment is also conducted. Mechanical properties of the base material were obtained by tensile tests. Residual stresses before and after the HFMI treatment were measured with the X-ray diffraction method. Hardness values on the weld toe were measured by the portable hardness tester. Stress concentration factors were calculated by the finite element analysis considering the geometry obtained by the 3D scanning. After all the already mentioned measurements, cyclic tests with constant amplitudes were performed until the failure of the welded details. These fatigue tests are necessary for the validation of the developed TSM model. The results of the developed and calibrated TSM model show good agreement with the fatigue test results. In order to expand the knowledge gained on the limited number of cyclic laboratory tests, a stochastic TSM model is used. The stochastic model generates a large number of "virtual tests" calibrated on a smaller number of results obtained by "real" laboratory tests mentioned earlier. In this way, the insights gained from a limited number of samples can be expanded to obtain a broader picture of a number of different stress amplitudes. The stochastic TSM model, by taking into account the uncertainty on the fatigue resistance side, gives more reliable results compared to the deterministic model, which in this case, means more reliable results.