Cold-formed steel (CFS) built-up columns are essential elements of CFS strap-braced shear walls, and understanding their cyclic behaviour is critical for the performance-based seismic design of such systems. Six CFS built-up columns were tested under cyclic axial loading employing the FEMA 461 protocol. The specimens comprised lipped (C-section) and unlipped (U-section) channels connected back-to-back with self-drilling fasteners at varying spacings, and were fixed to the testing machine using hold-downs representative of the connection detail used in practice. The cyclic response exhibited pinching, with strength and stiffness degradation following local–distortional interactive buckling in the C-sections and local buckling in the U-sections. Repeated cycles accumulated damage at the buckled corners of the web and flanges, leading to tensile tearing at these folds. Ultimate failure, however, was net-section rupture at the hold-down bolt holes during tension, influenced by out-of-plane shear lag and connection eccentricity at the web-bolted base. The Direct Strength Method accurately predicted the cyclic buckling strength of the C-sections but overestimated that of the U-sections. The Eurocode (EN 1993–1-3) net-section provision did not predict the observed tensile failure and overestimated the measured capacities, as it neglects shear lag, whereas the AISI S100 equation, which accounts for shear lag, was accurate for both section types. No damage was observed in the fasteners, and their spacing did not influence the cyclic behaviour. The C-sections dissipated 2.05 times more energy than the U-sections and, at the tensile peak-load cycle, developed an average equivalent viscous damping ratio of 0.27, compared with 0.20 for the U-sections.
