This article presents an exploratory multivariate statistical approach to gaining a more comprehensive understanding of variation in sub-aerial bone weathering in a British context. Weathering is one of the most common taphonomic modifications and provides a crucial line of evidence for reconstructing the taphonomic trajectories of faunal assemblages and archaeological deposits. It provides clear evidence for prolonged subaerial exposure, either prior to deposition in a context or as a result of later disturbance. In combination with other taphonomic indices such as gnawing, trampling, abrasion and fracture patterns, weathering can be used to reconstruct depositional histories and to investigate the structured treatment of different body parts or taxa. As a broad range of factors affect the prevalence and severity of weathering, weathering patterns can rarely be interpreted at face value indeed many variables such as predepositional microenvironment cannot be traced archaeologically. Other contributory factors pertaining to the structural properties of elements and taxa can be discerned and must be taken into account in interpreting weathering signatures although disagreement exists regarding which variables are most important in mediating weathering. In addition, for zooarchaeologists to interpret modification patterns, it is necessary for elements and taxa that are most likely to be affected by weathering to be defined. Deposits that are dominated by susceptible classes of remains are likely to exhibit greater modification than deposits that are not, even if depositional histories were similar. Through a combination of classification tree and ordinal regression analysis, this article identifies which archaeologically recoverable variables explain the greatest variance in weathering and which anatomical elements and taxa are most likely to be affected in archaeological deposits in the UK. Copyright © 2011 John Wiley & Sons, Ltd.