Evidence is presented that fibronectin is present on the platelet cell membrane and that it is a receptor for collagen in the platelet-collagen interaction. First, sodium dodecyl sulfate/acrylamide gel electrophoresis was performed on the proteins remaining attached to the surface of collagen after the removal of the remainder of the platelet by sonication. The material was relatively enriched in a glycoprotein that comigrated with cold-insoluble globulin (CIG), a form of fibronectin, and in other proteins which comigrated with myosin, actin, and tropomyosin. The presumptive presence of contractile proteins is consistent with the presence of microfibrillar proteins. Second, the collagen-attached material was shown to contain a protein that reacted'with anti-GIG serum by immunoelectrophoresis. Third, when CIG was preincubated with fibrous collagen, the platelet-collagen interaction was inhibited. Fourth, rabbit anti-human CIG stimulated human platelets to secrete the contents of their dense granules. The stimulation was not due to antibody complexes present in the solution. Fifth, a protein was extracted from wel -washed platelets and purified on affinity columns of anti-GIG antibodies. The isolated protein was found to bind to fibrous collagen. When the endothelial lining of blood vessels is breached, blood platelets come in contact with the components of connective tissue. The platelets adhere to the collagen in the connective tissue and are stimulated to release the contents of their dense granules, including ADP, Ca2+, and serotonin (1, 2). The Ca2+ and ADP are required for the subsequent platelet-platelet interactions, which lead to the formation of aggregates on a nucleus of the platelets bound to collagen (3). The aggregates form a platelet plug as an initial step in hemostasis. Because the collagen-platelet interaction is an important event in hemostasis, a great deal of interest has centered on the determination of the mechanism of the interaction. A major stimulus to interest in the collagen-platelet interaction came as a result of the investigations of Barber and Jamieson (4, 5). They presented evidence that membrane-bound glucosyl transferases on the surface of platelets bind to the galactose and glucose residues of collagen in an enzyme-substrate and enzyme-product complex, respectively. Further investigations by others tended to support this hypothesis (6, 7). However, evidence to the contrary has been presented. Menashi et al. (8) have shown that only denatured collagen can act as a substrate for the platelet glucosyl transferase, whereas only native, fibrous collagen serves as a platelet substrate in the collagen-platelet interaction (9, 10). In addition, others have shown that the carbohydrate residues of collagen can be completely destroyed by periodate oxidation without affecting the ability of the fibrous form of the oxidized collagen to interact normally with platelets (10, 11).Because the only current hypothesis as to the nature of the collagen receptor of platelets has been severe...