Atypical Equine Thrombasthenia (AET) is a recessive heritable platelet disorder in horses due to aberrant platelet signaling after thrombin stimulation, preventing platelets from fully activating or efficiently binding to fibrinogen, leading to prolonged bleeding. To date, AET has only been identified in the Thoroughbred horse breed. A prevalence study performed at one breeding farm found that one in every 150 Thoroughbreds was affected. Despite the negative effects on horse health and racing performance, the underlying etiology of this unique platelet disorder is unknown. Here we show that a variant in SEL1L, a gene which encodes a protein known to play a pivotal role in endoplasmic reticulum-associated protein degradation (ERAD) and not previously known to be involved in platelet function, leads to AET. In order to identify the underlying genetic mechanism, we performed whole genome sequencing on five affected, one equivocal, and eleven control Thoroughbred horses. A whole-genome association study was conducted, and associated variants were filtered by determining if they were in or near genes known to be expressed in platelets. This evaluation identified three putative loci: a missense variant in SEL1L (p.Ile604Val), a 1.6kb deletion in a long non-coding RNA ( AL355838.1) and a 73bp intronic deletion in VIPAS39. The long non-coding RNA AL355838.1 did not show expression in horse platelets, and there was no differential expression of SEL1L or VIPAS39 mRNA identified between affected and unaffected individuals. However, a significant difference in SEL1L protein expression was observed in AET-affected horses' platelets. Flow cytometry and immunofluorescence studies demonstrated that SEL1L is located intracellularly in equine platelets, but not in the α-granule, and moves to the surface of the platelet upon activation with thrombin. Although the template bleeding time was normal, platelets from horses that were homozygous for the SEL1L variant showed defective spreading on collagen. To support SEL1L as the causative gene and dissect the underlying mechanisms, we undertook studies in three additional models. Differentiation of human megakaryocytes revealed the presence of two SEL1L protein isoforms, p100 and p38, and both showed increased expression during megakaryopoiesis, although only p100 was delivered to mature platelets. We perfused whole blood from Mx1-Cre +; Sel1l fl/fl conditional knockouts and controls through a collagen-coated microfluidic chamber. We found that significantly fewer platelets from knockout mice bound to the collagen as compared to control siblings. Finally, a knockout zebrafish line was developed by deleting exons 3-20 using CRISPR/Cas9 mediated genome editing. Similar to the mouse, laser-mediated arterial endothelial injury of 5-day old mutant larvae resulted in significantly fewer thrombocytes (the fish platelet equivalent) adhering to the injured vessel wall as compared to wild-type siblings. Overall, these data support the conclusion that the SEL1L missense variant leads to AET, and that SEL1L has a previously undescribed and conserved role in platelet function, particularly the ability of platelets/thrombocytes to adhere to exposed collagen at sites of endothelial injury. These results will allow the Thoroughbred industry to easily diagnose AET-affected horses and make appropriate decisions for their racing and breeding careers. These data also provide an excellent introduction to further investigation into the understudied and potential role of ERAD proteins in human platelet disorders.