When designing according to the Eurocode, structures, that are not especially sensitive to non-linear behavior, are comprised by an individual member check. This member check includes a reduction factor to account for imperfections. The reduction factor is primarily dependent on the critical load, and this is why the determination of K-factors for individual member becomes important. Traditionally the individual K-factors are determined by methodical approaches such as the isolated subassembly or story methods. These methods do not take the actual system behavior into account, as the linear buckling analysis does. The system buckling approach, which uses the linear buckling analysis, is preferable when the structural behavior is complex, however the method contains the paradox, that compressive members with an relatively small axial force yields excessively large K-factors. In the present thesis a method to circumvent this paradox is proposed. The method distinguishes between members being prone to buckling and members that are not sensitive to additional axial load. Based on energy considerations the load multiplier found by the linear buckling analysis is weighed for each member. The proposed method is verified by 2D examples, where comparisons to known methods are made. It is found that the proposed method provides reasonable K-factors, and it is believed, that future work on the method could lead to an implementation in software, which provides automated code check. Thereby the burdensome manually definition of critical length for each members is avoided.