In this research, we explored the potential to reduce the cost of floating wind farms by adopting an integrated approach to optimally size semi-submersible substructures accounting for materials, fabrication and installation-logistics-related costs. A trade-off between manufacturing and installation costs was identified. This trade-off is driven by the growth of shipyard costs when the size of the structure increases, counteracting the reduction of fabrication costs achieved with a larger semi-submersible footprint. For the reference scenario, accounting for this trade-off yields a design that is a few tenths of a percent cheaper than when minimising only fabrication costs. However, the obtained design has a considerably smaller footprint than the fabrication-only case. The sensitivity of this trade-off to different installation strategies affecting the required storage area at the shipyard was assessed. When fabrication costs are dominant, the advantage of accounting for installation costs in the design process is negligible. Instead, larger storage area requirements increase the cost reduction achieved by optimising the semi-submersible while simultaneously accounting for fabrication and installation costs. The coupling effect remained significant for all the cases considered in a further sensitivity analysis of key parameters affecting the cost-optimal design. Furthermore, we identified several different designs that provide enough hydrostatic restoring moment in pitch to counteract the thrust-induced overturning moment within a small cost range from the most cost-effective one. This result suggests that additional criteria than minimising manufacturing and installation costs could drive the final design choice. @en