The Nucleosome Remodeling and Deacetylase complex (NuRD complex), a chromatin remodeller, is involved in transcriptional gene regulation. Thus far, the NuRD complex has been refractory to biochemical and structural analysis. This is, in part, due to difficulties with producing the complex recombinantly, as well as the low abundance and dynamic nature of the complex. As a result, a consensus for even fundamental parameters such as the subunit stoichiometry of the complex has not been reached. Here, we investigate both native whole complex as well as recombinantly produced subcomplexes. Using biochemical, proteomics, and structural biology approaches, we determine a stoichiometry of 4:2:2:1:1:1 for RBBP:HDAC:MTA:MBD:GATAD:CHD subunits, respectively. To our knowledge, for the first time we suggest that the stoichiometry of its core components is preserved across three cell lines that reflect three different human tissues. While the core subunits composition is conserved, some clear differences is observed in paralogue usage. Most importantly, only MBD3-NuRD (and not MBD2-NuRD) was observed in NTERA2 cell line. We also report dozens of potential partners for the complex. Of these, we demonstrate the direct interaction of two proteins namely, ZNF219 and SLC25A5. We also demonstrate that GATAD2A/B subunit is responsible for bringing the ATP-dependent chromatin remodelling activity (CHD4 subunit) into the deacetylase core of the complex and inhibiting the binding of a second MBD2/3 protein, leading to formation of an asymmetric complex.