Molecular chaperones are abundant cellular proteins that promote the folding, function and macromolecular assembly of a diverse set of substrate proteins, also known as clients. Regulation of chaperone specificity and function involves association with a growing set of distinct cofactors, or cochaperones. How these client, co-chaperone and chaperone interactions in vivo underpin the proteostasis network during development and disease remains unclear. We have combined mouse genetics, imaging and quantitative proteomics to systematically characterize a chaperone-cochaperone-client interaction network in mammalian motile ciliated cells. We uncover ZMYND10 to act as a specificity factor for the immunophilin cochaperone FKBP8 and HSP90 chaperone for the biosynthesis of the heavy chains of axonemal dyneins required for cilial beat. We establish a series of protein-protein interactions during cytoplasmic pre-assembly of macromolecular dynein motors, some of which are directly ZMYND10-dependent. In the absence of ZMYND10, failure of these interactions results in misfolded and unstable dynein heavy chains, which together with non-functional folding intermediates, are targeted for degradation. We propose that the motile ciliopathy Primary Ciliary Dyskinesia (PCD), such as seen in Zmynd10 mutants, be reconsidered as a disorder of protein folding, which opens new avenues for development of therapeutics.