Abstract Existing quantum processors contain far fewer qubits than the millions required to ensure that their economic value exceeds their cost, the point known as utility scale. Scaling these systems will involve surmounting several engineering challenges, such as the large-scale co-integration of qubits with low-power control electronics and the management of device variability. Many of these challenges have already been addressed by the complementary metal–oxide–semiconductor (CMOS) industry, which makes semiconductor spin qubits particularly promising for utility-scale quantum computing because of their inherent compatibility with industry processes. In this Review, we discuss the overlap between state-of-the-art semiconductor spin-qubit systems and the very-large-scale integration principles of the CMOS industry, identifying the main differences in terms of operation, materials and system requirements to bring spin-qubit systems to CMOS foundries. This relationship stands in contrast to other qubit systems that are being retrofitted for CMOS compatibility. We show that close collaboration between spin-qubit experts and their CMOS industry partners will accelerate the industrial-scale production of fault-tolerant processors. Key points - The quantum computing industry has begun to recognize that existing semiconductor manufacturing processes will be essential for scaling quantum devices into full-scale quantum computers containing millions of qubits. - The many modalities being explored as candidate qubits show varying degrees of compatibility with complementary metal–oxide–semiconductor (CMOS) technologies, the long-established industry standard. - Semiconductor spin qubits align naturally with CMOS processes, owing to their submicrometre footprint, planar layouts and electrostatic control. - However, some spin qubits impose requirements that diverge from standard CMOS practice, including millikelvin operation and unconventional material stacks. - Building quantum computers that provide true commercial value will require close co-design between the quantum and semiconductor industries, aligning fabrication, design rules and system architectures, rather than retrofitting one to the other. This is a preview of subscription content, access via your institution