The association of a rich lithic assemblage with a Middle Devensian mammalian assemblage at Lynford was initially thought indicative of a mammoth butchery locality, a rare occurrence for a European Middle Palaeolithic open site. However, taphonomic analyses suggest that the specimens have very different depositional histories and were incorporated into a palaeochannel in several stages. Most specimens are extremely fragmentary, probably the result of extensive trampling, and signs of weathering and root-damage provide further indications of exposure before burial. Carnivore damage is minimal but establishing the degree of interaction between the mammal fauna and alternative predators, such as Neanderthals, is problematic. Direct evidence of butchery is not present and the best indication of any form of mammoth exploitation lies in more circumstantial evidence such as the virtual absence of long bones from the main channel deposit and the mammoth age profiles. Instances of pathologies are also unusually common in the mammoths, implying that their greater vulnerability may have led to an accelerated demise either naturally or at the hands of a predator. The best evidence for direct faunal exploitation at the site is from green bone fractures and broken teeth that suggest marrow extraction in horse, reindeer and woolly rhinoceros.