Floral scent, often a complex mixture of several volatile organic compounds ( VOCs), has generally been interpreted as an adaptation to attract pollinators. However, not many studies have analysed which VOCs are functionally relevant for the reproductive success of a plant. Here, we show that, in Salix caprea ( Salicaceae), temporal changes in floral scent emission during the day and night attract two different types of flower visitor: bees during the day and moths during the evening and night. We analysed the contribution of the two flower visitor groups to the reproductive success of the plant. The differences in scent emitted during the peak activity times of flower visitors (day versus night) were quantified and the response of 13 diurnal/nocturnal pollinator taxa to the floral scents was tested using gas chromatographic and electroantennographic techniques. Many of the c. 40 identified scent compounds were physiologically active, and bees and moths responded to nearly identical sets of compounds, although the response strengths differed. In bioassays, bees preferred the most abundant 1,4-dimethoxybenzene over lilac aldehyde, a compound with increased emission at night, whereas moths preferred lilac aldehyde over 1,4-dimethoxybenzene. Pollination by wind plus nocturnal pollinators (mainly moths) or by wind alone contributed less to seed set than pollination by wind plus diurnal pollinators (mainly bees). This suggests that the emission of scent during the night and attracting moths have no significant effect on reproductive success. It is possible that the emission of lilac aldehydes and other compounds at night is s result of phylogenetic constraints. Future studies should investigate whether moths may produce a marginal fitness gain in some years and/or some populations. © 2014 The Linnean Society of London, Botanical Journal of the Linnean Society, 2014, 175, 624-640. [ABSTRACT FROM AUTHOR]