Sediment selection by settling larvae of the opportunistic polychaete Capitella sp. I was determined in laboratory still-water and flume experiments, where larvae were given a choice between two highly contrasting sediment treatments. In most cases, 2-h experiments were conducted with a natural, organic-rich mud and an abiotic, glass-bead mixture with a grain-size distribution similar to the mud, as the sediment treatments. Spatial settlement patterns were also determined in sediment arrays containing mud only. Two types of flume flows were tested, both with a near-surface velocity of ∼5 cm s−1, but one flow was cyclical, varying between about 2 and 7 cm s−1 with a period of 6.3 min, and one was steady with a boundary shear velocity of 0.26 cm s−1. Plastic spheres were added to the experiments as passive larval mimics. Capitella sp. I larvae selected the muddy sediment as opposed to the glass beads in all experiments conducted, consistent with food requirements of the deposit-feeding adults and with field distributions. Selectivity was insensitive to a range of experimental conditions, including flow, water temperature, light regime, experimental duration, distance sediment treatments were separated, time of year and larval batch. Experiments furthermore suggested that contact with the sediment is required to elicit a settlement response. Flows tested were weak compared to the range likely encountered by larvae of this species, even in depositional areas in the field; however, horizontal flow speeds within larval search distances of the bottom exceeded horizontal swim speeds of the larvae (determined in still water). A model for sediment selection in the field is proposed where larvae move up and down close to the bottom, while being transported by the flow, and test sediments on contact. Selection is thus accomplished by active acceptance or rejection of touchdown sites. This model was qualitatively supported by observations of larvae in still water and manipulative flume experiments. These results suggest that active sediment selection may be responsible, at least in part, for field distributions of this species.