Sex pheromones are chemical signals that mediate communication between males and females of a species and are important for mating. They have been well-studied in female moths, where females release volatile compounds to attract males over long distances. Butterflies fly during the day and were thought to rely more on vision. However, male pheromones in butterflies have also been described but have rarely been studied to the same detail as the female pheromones of night-flying moths. The Neotropical butterflies, Heliconius, have a long history of ecological and genetic studies. Mate choice trials have focused on male preference for female colour pattern. In addition, two types of pheromone have been described in Heliconius. The first is produced by a region of the wing known as the androconia, and the second is anti-aphrodisiac, transferred from male genitals to females during mating to prevent re-mating by the female. In this thesis I investigate the role of chemical signalling in mate choice of Heliconius. I describe sexual dimorphism in wing scale morphology and androconial chemical profile of H. melpomene. I demonstrate the importance of male chemical signalling for female choice in this and other Heliconius species. Male pheromones in other groups can convey information about mate quality. I show that larval but not adult diet affects the production of only minor components of the male pheromone. This suggests that male signalling cannot convey information about adult diet quality but could reflect larval diet. I survey variation in pheromone components across the broad geographic range of several species. I show that this provides a useful approach to identify biologically active components, which we expect to be species-specific over a large geographic range. There is also considerable interest in understanding the genetic basis for adaptive traits, and pheromones provide a tractable system for such studies. I uncover the genetic basis for anti-aphrodisiac production in H. melpomene through genetic mapping and use in-vitro assays for enzymatic activity to identify a novel terpene synthase activity in a gene family not previously known for this role. This finding suggests that terpene synthesis has evolved independently multiple times in insects. The work presented in this thesis is one of the first studies integrating both chemical ecology and genetics in butterflies. By combining behavioural, genetic and chemical studies, we can understand not only the composition of the chemical profile of individuals and species, but also their function and evolution.