Flowers, the reproductive structures of the approximately 400 000 extant species of flowering plants, exist in a tremendous range of forms and sizes, mainly due to developmental differences involving the number, arrangement, size and form of the floral organs of which they consist. However, this tremendous diversity is underpinned by a surprisingly robust basic floral structure in which a central group of carpels forms on an axis of determinate growth, almost invariably surrounded by two successive zones containing stamens and perianth organs, respectively. Over the last 25 years, remarkable progress has been achieved in describing the molecular mechanisms that control almost all aspects of flower development, from the phase change that initiates flowering to the final production of fruits and seeds. However, this work has been performed almost exclusively in a small number of eudicot model species, chief among which is Arabidopsis thaliana. Studies of flower development must now be extended to a much wider phylogenetic range of flowering plants and, indeed, to their closest living relatives, the gymnosperms. Studies of further, more wide-ranging models should provide insights that, for various reasons, cannot be obtained by studying the major existing models alone. The use of further models should also help to explain how the first flowering plants evolved from an unknown, although presumably gymnosperm-like ancestor, and rapidly diversified to become the largest major plant group and to dominate the terrestrial flora. The benefits for society of a thorough understanding of flower development are self-evident, as human life depends to a large extent on flowering plants and on the fruits and seeds they produce. In this preface to the Special Issue, we introduce eleven articles on flower development, representing work in both established and further models, including gymnosperms. We also present some of our own views on current trends and future directions of the flower development field.