The SEL1L-HRD1 protein complex of endoplasmic reticulum (ER)-associated degradation (ERAD) plays indispensable roles for many physiological processes in a substrate-specific manner; however, the nature of endogenous substrates remains largely elusive. Here we have developed a unique proteomics strategy based on the intrinsic property of the SEL1L-HRD1 ERAD complex to identify potential endogenous ERAD substrates in human kidney cell line HEK293T and mouse brown adipocytes. Over 100 potential substrates involved in many cellular processes, including both membrane and luminal proteins regardless of their glycosylation and disulfide bond status, are identified in each cell type, among which 34 are shared. We further uncover SEL1L-HRD1 ERAD as a suppressor of the biogenesis of glycosylphosphatidylinositol (GPI)-anchored proteins via degrading a key subunit of the GPI-transamidase complex known as phosphatidylinositol glycan anchor biosynthesis class K protein (PIGK). Lastly, several PIGK disease variants are highly unstable and quickly degraded by SEL1L-HRD1 ERAD. This study shows the most effective way to identify cell type-specific proteome-wide potential endogenous SEL1L-HRD1 substrates, and uncovers a new function of SEL1L-HRD1 ERAD in the biogenesis and disease pathogenesis associated with GPI-anchored proteins