The design of compression members typically consists of assigning appropriate effective length coefficients. This is a method used to determine relative slenderness and then – capacity. For tubular elements, various guidelines, such as by CIDECT, are provided. However, a different approach proves to be more accurate. Capacity is evaluated by axial stress reduction function, and then by Linear Buckling Analysis (LBA). European buckling curves are used for stiffness reduction function, relying on their ability to capture accurately the response of real members. In a truss system, the interaction between members through their relative stiffnesses is therefore captured. Something that effective length coefficients cannot do, especially in the inelastic stage, with imperfections and residual stresses. A comparative study is performed for typical truss structures with tubular elements. This is done to evaluate areas where the effective length approach underestimates or overestimates capacity. The study aims to provide a more accurate approach, leading to safer and more economical design of truss members. Linear buckling analysis (LBA), with joint stiffness included, is performed for the reduced stiffness method. Joint rigidity is evaluated through Finite Element analysis in ANSYS software. The material model in ANSYS was calibrated against laboratory-tested tensile coupons. The influence of joint stiffness on member's capacity is assessed. Members generally display greater capacity than in the case of the effective length method. Differences can reach 25%. Proposed method is more intuitive, and better represents the real structure’s behaviour. This provides more confidence for the designer. The need for effective length coefficients is thus removed, and individual checks for each member are no longer required. Load multiplier greater than 1.0 becomes the design criterion for compression members of a truss with little bending. This approach can also be used in the assessment of existing structures.