The ubiquitin receptors RPN10 and RPN13 harbor multiple activities including ubiquitin binding; however, solid evidence connecting a particular activity to specific in vivo functions is scarce. Through complementation, the ubiquitin-binding site-truncated Arabidopsis RPN10 (N215) rescued the growth defects of rpn10-2, supporting the idea that the ubiquitin-binding ability of RPN10 is dispensable and N215, which harbors a vWA domain, is fully functional. Instead, a structural role played by RPN10 in the 26S proteasomes is likely vital in vivo. A site-specific variant, RPN10-11A, that likely has a destabilized vWA domain could partially rescue the rpn10-2 growth defects and is not integrated into 26S proteasomes. Native polyacrylamide gel electrophoresis and mass spectrometry with rpn10-2 26S proteasomes showed that the loss of RPN10 reduced the abundance of double-capped proteasomes, induced the integration of specific subunit paralogues, and increased the association of ECM29, a well-known factor critical for quality checkpoints by binding and inhibiting aberrant proteasomes. Extensive Y2H and GST-pulldown analyses identified RPN2-binding residues on RPN13 that overlapped with ubiquitin-binding and UCH2-binding sites in the RPN13 C-terminus (246–254). Interestingly, an analysis of homozygous rpn10-2 segregation in a rpn13-1 background harboring RPN13 variants defective for ubiquitin binding and/or RPN2 binding supports the criticality of the RPN13–RPN2 association in vivo.