The crown-of-thorns starfish (COTS, the Acanthaster planci species group) is a highly fecund predator of reef-building corals throughout the Indo-Pacific region 1 . COTS population outbreaks cause substantial loss of coral cover, diminishing the integrity and resilience of reef ecosystems 2-6 . Here we sequenced genomes of COTS from the Great Barrier Reef, Australia and Okinawa, Japan to identify gene products that underlie species-specific communication and could potentially be used in biocontrol strategies. We focused on water-borne chemical plumes released from aggregating COTS, which make the normally sedentary starfish become highly active. Peptide sequences detected in these plumes by mass spectrometry are encoded in the COTS genome and expressed in external tissues. The exoproteome released by aggregating COTS consists largely of signalling factors and hydrolytic enzymes, and includes an expanded and rapidly evolving set of starfish-specific ependymin-related proteins. These secreted proteins may be detected by members of a large family of olfactory-receptor-like G-protein-coupled receptors that are expressed externally, sometimes in a sex-specific manner. This study provides insights into COTS-specific communication that may guide the generation of peptide mimetics for use on reefs with COTS outbreaks.COTS are extremely fecund mass spawners 7 , which predisposes them to population outbreaks that result in a pronounced loss of live coral cover and associated biodiversity. These outbreaks have a higher impact on reef health and resilience than the combined effects of coral bleaching and disease, and increase the susceptibility of reefs to other potentially detrimental events, such as severe storms [2][3][4][5][6] (Supplementary Note 1).Although a range of local in situ control measures have been applied with some success (Supplementary Note 1), mitigation of COTS outbreaks on the necessary regional scale requires mass-deployed, species-specific strategies. In this context, genome-encoded COTSspecific attractants that underpin spawning aggregations have substantial potential as biocontrol agents. To identify attractants, we sequenced the genomes of two wild-caught individuals separated by over 5,000 km, one from the Great Barrier Reef (GBR), Australia and the other from Okinawa (OKI), Japan (Fig. 1c, d and Extended Data Fig. 1). We also sequenced transcriptomes from external organs, and proteins released into the seawater by COTS that were aggregating or were in the presence of their main predator, the giant triton Charonia tritonis (Fig. 1b).We generated separate 384 megabase (Mb) draft assemblies for the GBR and OKI genomes (Extended Data COTS genes are labelled and are marked with red lines; other asteroids, two shades of orange and yellow lines; sea urchins, dark green; hemichordates, light green; molluscs, pink; annelids, purple; cnidarians, black; and vertebrates, blue. The three clades to which COTS sequences belong are indicated by the outer circle. The asterisk denotes the fish-specific tru...