THE occurrence of small amounts of proinsulin in crystalline preparations of bovine, porcine, and rat insulin1–4 and the many similarities in properties of proinsulin and insulin prompted us to consider the possibility that proinsulin co-crystallizes with insulin. Spectral5 as well as immunological6 studies have provided evidence that the insulin moiety in proinsulin has a very similar conformation to that of insulin, and proinsulin exhibits the same tendency to form dimers in acidic solution and higher polymers, mainly hexamers, in neutral solutions containing zinc ions7. Studies on crystals of proinsulin by Low and coworkers8 have confirmed the presence of dimers in the unit cells of these crystals. All these observations suggest that the connecting polypeptide segment in proinsulin is oriented away from those surfaces involved in dimer and hexamer formation.