Lizards, snakes and amphisbaenians together constitute the Squamata, the largest and most diverse group of living reptiles. Despite their current success, the early squamate fossil record is extremely patchy. The last major survey of squamate palaeontology and evolution was published 20 years ago. Since then, there have been major changes in systematic theory and methodology, as well as a steady trickle of new fossil finds. This review examines our current understanding of the first 150 million years of squamate evolution in the light of the new data and changing ideas. Contrary to previous reports, no squamate fossils are currently documented before the Jurassic. Nonetheless, indirect evidence predicts that squamates had evolved by at least the middle Triassic, and had diversified into existing major lineages before the end of this period. There is thus a major gap in the squamate record at a time when key morphological features were evolving. With the exception of fragmentary remains from Africa and India, Jurassic squamates are known only from localities in northern continents (Laurasia). The situation improves in the Early Cretaceous, but the southern (Gondwanan) record remains extremely poor. This constrains palaeobiogeographic discussion and makes it difficult to predict centres of origin for major squamate clades on the basis of fossil evidence alone. Preliminary mapping of morphological characters onto a consensus tree demonstrates stages in the sequence of acquisition for some characters of the skull and postcranial skeleton, but many crucial stages--most notably those relating to the acquisition of squamate skull kinesis--remain unclear.