Marine pelagic larvae from throughout the animal kingdom use a hierarchy of environmental cues to identify a suitable benthic habitat on which to settle and metamorphose into the reproductive phase of the life cycle. The majority of larvae are induced to settle by biochemical cues and many species have long been known to preferentially settle in the dark. Combined, these data suggest that larval responses to light and biochemical cues may be linked, but this is yet to be explored at the molecular level. Here, we track vertical position of larvae of the sponge Amphimedon queenslandica to show that they descend to the benthos at twilight, by which time they are competent to respond to biochemical cues, consistent with them naturally settling in the dark. We then conduct larval settlement assays under three different light regimes (natural day-night, constant dark or constant light), and use transcriptomics on individual larvae to identify candidate molecular pathways underlying the different settlement responses that we observe. We find that constant light prevents larval settlement in response to biochemical cues, likely via actively repressing chemostransduction this is consistent with the sustained upregulation of a photosensory cryptochrome and two putative inactivators of G-protein signalling in the constant light only. We hypothesise that photo- and chemosensory systems may be hierarchically integrated into ontogeny to regulate larval settlement via nitric oxide (NO) and cyclic guanosine monophosphate (cGMP) signalling in this sponge that belongs to one of the earliest branching of the extant animal lineages. Overall design: Transcriptomic data for 43 individual larvae of the sponge Amphimedon queenslandica aged 1 and 9 hours post emergence (hpe) that were exposed to one of three experimental light regimes. Larvae were collected within a 30-minute window beginning at 11 am and were distributed into 0.22 µm filtered sea water, under one of three experimental light regimes, by 0.5 hours post emergence (hpe). The light regimes were (1) natural day-night, (2) constant dark (using LEE filter model 299, 1.2 neutral density, allowing 6.6% transmission of ambient daylight), and (3) constant light (full spectrum Lumilux T5 HE tube). For RNA-sequencing, two timepoints (1 and 9 hpe) were chosen to represent precompetent and competent larvae, respectively.