Biaryls are ubiquitous core structures in drugs, agrochemicals, and organic materials which have profoundly improved many aspects of our society. Although traditional cross-couplings have made practical the synthesis of many biaryls, C–H arylation represents a more attractive and cost-effective strategy for building these structural motifs. Furthermore, the ability to install biaryl units in complex molecules via late-stage C–H arylation would allow access to valuable structural diversity, novel chemical space and intellectual property in only one step. However, known C–H arylation protocols are not suitable for substrates decorated with polar and delicate functionalities, which are commonly found in molecules possessing biological activity. Here we introduce a class of ruthenium catalysts displaying unique efficacy towards late-stage arylation of heavily-functionalized substrates. The design and development of this class of catalysts was enabled by a mechanistic breakthrough on the Ru(II)-catalysed C–H arylation of N–chelating substrates with aryl (pseudo)halides, which has remained poorly understood for nearly two decades.