One of the most dynamic phases of vertebrate evolutionary history, characterized by two contrasting patterns of events, took place in the Cretaceous (165-145 Ma). On the one hand, many ‘classic’ Mesozoic groups such as ornithischian and sauropod dinosaurs, pterosaurs, and marine reptiles, reached peak levels of taxonomic, morphological and ecological diversity, but subsequently declined and became extinct, either before the end of the Cretaceous, or in the event that punctuated the end of this period. On the other hand several modern groups of vertebrates, including the teleosts, turtles, squamates (lizards and snakes), birds and mammals, radiated during this interval and although some of these suffered a drop in diversity at the K-T boundary, none of them became extinct. Not all groups fit neatly within these patterns, however. The choristoderes, for example, a group of gavial-like aquatic reptiles, have a long Mesozoic record, survived the K-T event, and only finally became extinct in the Oligocene (Evans and Hecht 1993). Patterns of vertebrate evolution in the Cretaceous have been derived from three main sources of evidence: first, and most importantly, the fossil record; second, phylogenetic analyses based on the anatomy of extinct and extant taxa; and third, molecular phylogenies, although so far, these have been almost completely restricted to modern taxa. Taking these three categories in reverse order, molecular phylogenies are playing an increasingly important réle in the reconstruction of the early history of major living groups such as birds and mammals, but this work has often been highly controversial. As an example, molecular phylogenies proposed by Sibley and Ahlquist (1990) and, more recently, by Hedges et al. (1996) suggested that early representatives of most if not all orders of living birds had already appeared before the end of the Cretaceous, possibly much earlier. These claims have sparked considerable debate between molecular phylogeneticists and palaeontologists, although some rapprochement between the two groups now seems to be under way following recent revisions of estimates for the timing of cladogenetic events, and the discovery, in Late Cretaceous deposits, of fossil remains that seem to belong to several orders of neornithine birds (e.g. Hope 1997; Hutchison et al. 1997). Ultimately, however, the potential for molecular phylogenies to give insights into events that took place in the Cretaceous is likely to prove somewhat limited, since this methodology cannot inform us as to the existence or history of extinct lineages, the total number of which is likely to exceed considerably that of their living relatives. Turning to phylogenetics, previously palaeontologists reconstructed evolutionary histories essentially by aligning fossil taxa from different stratigraphical horizons into ancestor—descendant sequences. The widespread adoption of phylogenetic systematics (cladistics) during the last 20 years has had a radical impact both on the reconstruction of patterns and the way in which they are interpreted. To give a simple, well known example, this method enables the correct elucidation of potentially misleading situations where taxa nearer the crown group occur in older strata than those further from the crown group, or vice versa. Perhaps more importantly, cladistic studies can be used to predict the existence of ‘ghost lineages’ as yet unknown from fossils. This can have a profound impact on our understanding of the history of major clades: for example, a recent study of the relationships of sauropod dinosaurs (Wilson and Sereno 1998) suggested that many important lineages, as yet unknown in the fossil record, existed in the Early and Mid Jurassic. Whilst cladistics and molecular phylogenetics have much to offer, the fossil record remains the principal source of evidence regarding patterns of vertebrate evolution in the Cretaceous and, irrespective of their origin, the final arbiter of hypotheses regarding these patterns. Huge collections of fossil vertebrate material centuries, partly as a result palaeontologists. The 1990s from Cretaceous deposits have been accumulated over the last two of commercial activity, and partly through organized collecting by have witnessed a series of dramatic discoveries, sometimes from new localities such as the Early Cretaceous lake deposits of Liaoning, China, the source of numerous primitive birds and some non-avian feathered dinosaurs (Ji et al. 1998), or from long-known locations, for example the Late Cretaceous aeolian deposits of the southern Gobi desert, Mongolia, which recently yielded the remains of a theropod dinosaur, Oviraptor, preserved intact and lying upon a nest containing what are now presumed to be its eggs (Norrell et al. 1995). Despite these discoveries, the fossil record of Cretaceous vertebrates remains highly biased in at least two important ways. First, although the Late Cretaceous is relatively well represented, much of the rest of the record is rather poor. The Early Cretaceous is patchy, and the late Early to early Late Cretaceous interval is notoriously incomplete, especially for terrestrial vertebrates (e.g. Benton 1987). Second, the vast majority of discoveries, so far, has been made in the northern hemisphere. Thus, whilst North America and, to some extent, Eurasia are relatively well represented, the record for South America is very patchy and, with one or two notable exceptions, very little is yet known from Africa or Australia. The nine contributions that form this volume add to our knowledge of Cretaceous fossil vertebrates, but the important feature of these papers is that they report on taxa that have been recovered from stratigraphical intervals or palaeogeographical regions that have a relatively poor fossil record. Selachians (Underwood and Mitchell), a lizard (Evans and Barbadillo), a snake (Gardner and Cifelli), a choristodere (Evans and Manabe), goniopholidid crocodiles (Salisbury et al.) and ankylosaurian dinosaurs (Pereda Suberbiola and Barrett) reported in this volume were recovered from the Lower, or the lower Upper Cretaceous, and in most cases the genera described represent the first record for a particular stage. The snake described by Gardner and Cifelli is of particular interest as it also represents the oldest record for North American ophidians. In complementary fashion, the African teleost Enchodus (Cavin), the Japanese choristodere (Evans and Manabe), and the Madagascan crocodile (Buckley and Brochu) are all welcome additions to our knowledge of vertebrate faunas from regions that are as yet poorly represented. The new crocodile is a particularly timely discovery, since the origin and palaeobiogeographical history of Madagascan vertebrates has long been, and remains, a focus of interest and controversy. The crocodile, which forms part of an important new collection from the Upper Cretaceous of the Mahajanga Basin, adds further weight to the idea that faunal exchange across Gondwana continued until late in the Mesozoic, primarily via Antarctica, rather than directly between Africa and South America. The final contribution, an analysis of stance and gait in ornithopods based on fossil tracks from the Purbeck of England (Wright), is an excellent example of how another type of gap in our palaeobiological knowledge, direct evidence of the locomotory ability of fossil vertebrates, might be filled. Vertebrate palaeoichnology is currently undergoing a long-awaited renaissance and Wright’s study of the Purbeck tracks, which reveals the correct orientation and positioning of iguanodontid forelimbs as they walked, is typical of the work that is showing how important palaeobiological information can be wrested from this type of fossil material.