Palaeo-ornithology has been dominated by taxonomy. To try and redress the balance and help palaeoecologists interpret fossil birds in a biological and ecological perspective, the taphonomy of birds needs to be fully understood. The taphonomy of birds is concerned with all processes from death to the collection of the fossil bird. Between these two points (the transfer of the organism from the biosphere to the lithosphere) a variety of forces and processes affect the bird/fossil. By means of experiments in the natural environment and in controlled conditions in the laboratory, and subsequent comparisons of the results with case studies of fossil assemblages, the processes leading to preservation can be deduced and the former living community restored on the basis of the fossil evidence. The research involved two main approaches: 1. experimental taphonomy / observational taphonomy; and 2. case histories of fossil communities and their interpretation. Experimental work was carried out in the natural environment. Two field sites were chosen in southern Florida, a freshwater environment and a marine environment. The monitoring and controlling of these experiments required knowledge and techniques in zoology, botany, ecology, sedimentology, limnology, marine biology, microbiology, pathology and forensic science. Results obtained included the effects of scavenging, anoxia, transport, rate of burial, and temperature on rates of decay, the causes of bird mortality, the processes resulting in disarticulation, and the effects of decay upon feathers. Once the experimentaVobservational data had been collected they allowed a series of taphonomic thresholds (a decay sequence) to be defined. These data were then applied to case studies of fossil bird assemblages from different sedimentological environments. The following LagersHitten were investigated: Messel (Eocene, Germany) = restricted lacustrine; Green River (Eocene, USA) = lacustrine; Solnhofen Lithographic Limestone (Jurassic, Germany) = restricted marine; La Meseta Formation (Eocene, Antarctica) = marine; Rancho La Brea (Pleistocene, USA) = terrestrial "trap". The biases in each environment were assessed (e.g. birds in an aquatic ten-estrial environment had a higher preservation potential than birds from a tenestrial environment). The fossil record of birds is not as depauperate as previously thought but is heavily biased, depending on the proximity of the bird's habitat to that of the preserving sedimentary environment. Marine and littoral birds are poorly represented even though they inhabit sedimentary environments with a high preservation potential. This reflects low densities of birds per unit area. Aquatic birds (and terrestrial birds that inhabit the ecotone surrounding freshwater together with some larger fOlIDS from further away) are much better represented. This is because they inhabit the only terrestrial environments with a high preservation potential, coupled with the high densities of individuals per unit area. The bias towards large terrestrial birds is due to their large bones being more resistant to transport induced damage. These results have implications for the understanding of the evolution of birds. Patterns of evolution in birds can not be fully resolved on fossil evidence alone; biases in the taphonomy of birds only permit a small proportion of species from certain environments to be preserved. The taphonomy of feathers was investigated and it was discovered that the "organic trace" that commonly represents the outline of the feather trace is the diagenetically altered glycocalyx of the bacteria that were degrading the feather. In several localities these feather-degrading bactelia are preserved in authigenic minerals. The taphonomy of bats and pterosaurs was also investigated. The similarity of anatomical structures of birds, bats and pterosaurs results in similar taphonomic pathways.