The Burgess Shale arthropod Leanchoilia superlata Walcott 1912, commonly preserves a three-dimensional axial structure generally interpreted as gut contents. Thin-section examination shows this instead to be phosphatized biserially repeated midgut glands, including exceptional preservation of subcellular features. The preferential mineralization of these structures is related to their unusually high chemical reactivity and probably to an internal source of phosphate. Sub-millimetric lineations previously interpreted as annular musculature are in fact planar, sometimes radially arranged, subdivisions of these glands. Ventral rows of isolated phosphate patches appear to represent the same tissue.In extant arthropods, extensively developed midgut glands are related to a rich but infrequent diet with a primary function in storage. Their conspicuous occurrence in unambiguous fossil predators such as Sidneyia and Laggania (Anomalocaris) suggests they served a similar role in the Cambrian; by extension, their conspicuous occurrence in Leanchoilia suggests it was a predator or scavenger.Phosphatized midguts with a structure essentially indistinguishable from that of Leanchoilia are also found in Burgess Shale Odaraia, Canadaspis, Perspicaris, Sidneyia, Anomalocaris, and Opabinia. All are characterized by a distinctive sub-millimetric arrangement of planar elements that is not found in extant arthropods or trilobites, suggesting they diverged before the last common ancestor of extant forms; i.e., they represent stem-group arthropods.Three-dimensionally preserved guts are widely preserved in the Lower Cambrian Chengjiang biota but, unlike those in the Burgess Shale, appear to be filled with sediment. Although generally interpreted as evidence of deposit feeding, the form of these structures points to early permineralization of (sediment-free) midgut glands that were subsequently altered to clay minerals. There is no evidence of deposit feeding in the Chengjiang; indeed, there is a case to be made for deposit feeding not being generally exploited generally until after the Cambrian.Fossils with three-dimensionally preserved axes from the Lower Cambrian Sirius Passet biota have been interpreted as lobopodians; however, most of the putative lobopodian features find alternative interpretations as aspects of Leanchoilia-type midgut glands. Although Kerygmachela is reliably identified as a stem-group arthropod, its phylogenetic position remains unresolved owing to the non-preservation of critical external features and to the plesiomorphic nature of its Leanchoilia-type midgut.