This paper reports the development, numerical implementation and application of a novel methodology to perform a multi-objective stiffening optimisation of conventional cold-formed steel (CFS) cross-section shapes with longitudinal intermediate stiffeners, manufactured by bending the steel sheets with roll-forming machines. The methodology is based on a Genetic Algorithm (GA) and aims at maximising the compressive and/or flexural ultimate strength of CFS members currently used in industry with respect to local, distortional and local-distortional interactive failures. The design approach adopted to estimate the member ultimate strengths is based on the Direct Strength Method (DSM) and Generalised Beam Theory (GBT) is used to calculate the elastic buckling loads and moments. The proposed methodology consists of adding small intermediate stiffeners, whose possible locations are pre-defined in accordance with a map of executable fold-line positions, thus leading to an optimisation procedure whose output are alternative (stiffened) cross-sections that (i) outperform the original conventional/plain cross-section and (ii) can be fabricated with a minimal additional cost – the original cross-section typology, overall dimensions and wall angles are retained. The application and potential of the developed GA-based code is illustrated by addressing the stiffening optimisation of conventional lipped channel, lipped zed and rack cross-section shapes for CFS members subjected to compression, bending or compression and bending – the numerical results are presented in terms of cross-section families belonging to the optimal Pareto Front. For validation and assessment purposes, some optimised cross-sections obtained are compared with the “exact” solutions provided by a “brute-force search”.