Recent behavioural experiments have shown that birds use ultraviolet (UV)–reflective and fluorescent plumage as cues in mate choice. It remains controversial, however, whether such UV signals play a special role in sexual communication, or whether they are part of general plumage coloration. We use a comparative approach to test for a general association between sexual signalling and either UV-reflective or fluorescent plumage. Among the species surveyed, 72% have UV colours and there is a significant positive association between UV reflectance and courtship displays. Among parrots (Psittaciformes), 68% of surveyed species have fluorescent plumage, and again there is a strong positive association between courtship displays and fluorescence. These associations are not artefacts of the plumage used in courtship displays, being generally more ‘colourful’ because there is no association between display and colours lacking UV reflectance or fluorescence. Equally, these associations are not phylogenetic artefacts because all results remain unchanged when families or genera, rather than species, are used as independent data points. We also find that, in parrots, fluorescent plumage is usually found adjacent to UV–reflective plumage. Using a simple visual model to examine one parrot, the budgerigar Melopsittacus undulatus, we show that the juxtaposition of UV–reflective and fluorescent plumage leads to a 25–fold increase in chromatic contrast to the budgerigar's visual system. Taken together, these results suggest that signals based on UV contrast are of special importance in the context of active sexual displays. We review briefly six hypotheses on why this may be the case: suitability for short–range signalling; high contrast with backgrounds; invisibility to predators; exploitation of pre–existing sensory biases; advertisement of feather structure; and amplification of behavioural signals.