Complex I (NADH:ubiquinone oxidoreductase) is the first enzyme of the mitochondrial respiratory chain and is composed of 45 subunits in humans, making it one of the largest known multi-subunit membrane protein complexes1. Complex I exists in supercomplex forms with respiratory chain complexes III and IV, which are together required for the generation of a transmembrane proton gradient used for the synthesis of ATP2. Complex I is also a major source of damaging reactive oxygen species and its dysfunction is associated with mitochondrial disease, Parkinson’s disease and ageing3,4,5. Bacterial and human complex I share 14 core subunits that are essential for enzymatic function; however, the role and necessity of the remaining 31 human accessory subunits is unclear1,6. The incorporation of accessory subunits into the complex increases the cellular energetic cost and has necessitated the involvement of numerous assembly factors for complex I biogenesis. Here we use gene editing to generate human knockout cell lines for each accessory subunit. We show that 25 subunits are strictly required for assembly of a functional complex and 1 subunit is essential for cell viability. Quantitative proteomic analysis of cell lines revealed that loss of each subunit affects the stability of other subunits residing in the same structural module. Analysis of proteomic changes after the loss of specific modules revealed that ATP5SL and DMAC1 are required for assembly of the distal portion of the complex I membrane arm. Our results demonstrate the broad importance of accessory subunits in the structure and function of human complex I. Coupling gene-editing technology with proteomics represents a powerful tool for dissecting large multi-subunit complexes and enables the study of complex dysfunction at a cellular level.