Piled friction soil show a set-up which is increasing with the time after installation. The cost could be reduced if the set-up had been taken into account when designing but in Sweden it is very unusual to do that. Previous studies showed that set-up can vary between 25 % and 75 % during the first 90 days after pile installation. This showed that the effect of the phenomenon is difficult to predict. A study involving five projects have been carried out by specifying certain parameters like pile type, method for pile installation and test method for measure the bearing capacity in friction soils located in Sweden. The expectations from this study were whishes that the large dispersion should be reduced. This was not the case. One example of the wide dispersion is two piles which are 29 meters long and installed in Södertälje. One of the piles had a set-up of 249 % for 21 days while the other one had a loss of 7 % in bearing capacity during the same time interval. Since these large dispersions encountered it was found that soil characteristics and composition has a major impact on the coming set-up. Friction soil exhibit variations in properties such as how dense it is, the variety and size of layers in the soil profile and grain size. These are all not possible to determine if it not is a research project or some kind of project where detailed surveys are can be done considering time and money. Another parameter that indicated to have impact on the outcome of the phenomenon was the order of installations of the piles. Stress rearrangement in the soil is one of the mechanisms behind the set-up. An empirical relationship to predict the bearing capacity has been used as the piles have been compared with each other. An interesting finding of the study showed that the pile length influences how large the pile set-up gets. One design method for piles that takes into account the pile set-up is the ICP-method but is barely used in Sweden. To see how well and if this method match the results from dynamic pile tests calculations has been done. Average showed that the ICP-method gave a 13 % lower bearing capacity than the real value. There are several reasons that this method differs from the values from test data. For example the CPT tests were not as deep as the piles so the stress from the pile toe is just extrapolated. The friction angle in the soil is based on empirical data which means that the friction angle between concrete and soil is empirical as well which affect the result significantly.