From a detailed analysis of published and unpublished sources, we constructed a digitized three-dimensional, stratigraphically-controlled excavation grid of Zhoukoudian Locality 1 in order to assess the spatial relationships of the excavated materials. All 15 fossil Homo erectus loci were mapped on the grid. Meter cubes were used in excavation starting in 1934, and Loci H through O, established between 1934 and 1937, were mapped to within 1 m(3)vertical and horizontal provenience. Loci A through G, established between 1921 and 1933, were excavated in the northernmost part of Locality 1 by unmapped quarrying, but their stratigraphic levels were recorded. We could localize Loci A through G on the grid system by utilizing locations of remaining walls, stratigraphic sections, excavation reports, excavation maps, and photographs. Loci contained skeletal elements of Homo erectus individuals scattered over areas of the cave floor of up to 9 m in diameter. Scoring of taphonomic damage on the Homo erectus sample, as observed on casts and originals, demonstrates that 67% of the hominid sample shows bite marks or other modifications ascribed to large mammalian carnivores, particularly the large Pleistocene cave hyena, Pachycrocuta brevirostris. Virtually all of the remaining Homo erectus skeletal assemblage shows breakage consistent with this taphonomic pattern of fragmentation. Bioturbation by digging carnivores is the most likely explanation for a fragment of Homo erectus Skull XI discovered 1 m below its other conjoined portions in Locus L. Carbon on all the Homo erectus fossils from Locus G, a circumscribed area of 1-meter diameter, earlier taken to indicate burning, cooking, and cannibalism, is here interpreted as detrital carbon deposited under water, perhaps the result of hyaenid caching behavior. Locus G records the close stratigraphic and horizontal association of stone artifacts with Homo erectus and other vertebrate skeletal elements, an association that is seen at other loci as well. Layer 4 of the excavation contains equid cranial bone previously interpreted to have been burned while fresh. We here document that Locus B Homo erectus, including Skull I, is stratigraphically associated with this evidence, but at some 10-12 m distance. Even though the presence of wood-stoked fires and hearths is not supported by geochemical results, evidence of fire at Locality 1 in the form of burned bone is confirmed. Contextual relationships of fossil skeletal elements, relationships of carnivore damage and stone tool cutmarks on bone, and evidence of the burning of fresh bone associated with Homo erectus and stone tools support a model of transient hominid scavenging aided by the use of fire at the large hyenid den that became Zhoukoudian Locality 1. Although the original excavation catalogue from Locality 1, as well as a significant number of fossils and stone artifacts, were lost during World War II, catalogue numbers on the many surviving specimens can be used to locate fossils and artifacts within the three-dimensional grid provided in this paper.