The distribution of horseshoe crab (Limulus polyphemus) larvae and early benthic juveniles is often non-random, with higher abundances typically found in areas of dense submerged vegetation (e.g., seagrass and algae) in coastal and estuarine nursery habitats. To determine if these distribution patterns are mediated by behavioral responses to chemical odors associated with potential settlement sites, we examined whether habitat chemical cues influence the swimming behavior (chemically-mediated rheotaxis) of horseshoe crab larvae at two different periods in their development: immediately after hatching (early-stage) and just prior to metamorphosis (late-stage). The hypothesis that larvae exhibit a positive rheotaxis (upstream movement) in the presence of chemical cues from submerged vegetation and negative rheotaxis (downstream movement) in response to odors from predators was tested in a laboratory T-maze. Larvae were exposed to seawater containing chemical odors from one of four sources: (1) water collected directly from the estuary, and odor water prepared with (2) aquatic vegetation, (3) potential predators, or (4) conspecifics. Behavioral responses of larvae varied with the source of the odor and the developmental stage of the larva. Both larval stages responded to estuarine water and water containing cues from the seagrass Halodule wrightii, a known nursery habitat, but response rates were greater for early-stage larvae. The response of larvae to H. wrightii odor was dose-dependent; upstream swimming response increased with increasing H. wrightii odor concentration. Though juvenile horseshoe crabs are also found within beds of Syringodium filiforme and Acanthophora spicifera, in our experiments early-stage larvae swam downstream away from both cues. In contrast, late-stage larvae exhibited a positive rheotactic response to the odor of A. spicifera. Conspecific odor had no effect on the rheotaxis of either developmental stage tested. Both larval stages unexpectedly responded to predator odors (blue crabs, Callinectes sapidus; killifish, Fundulus grandis) with a positive rheotaxis. When combined with a known positive cue (H. wrightii), killifish predator odor did not alter the rheotactic response of early-stage larvae; however, the addition of the predator odor reversed the upstream swimming behavior of late-stage larvae, likely indicating an alarm response to the presence of the predatory fish. Collectively, our results indicate that the distribution patterns of early benthic horseshoe crabs are a consequence of developmental shifts in oriented swimming behaviors (rheotaxes) that are mediated by chemical odors associated with settlement habitats. These behaviors should aid in the larva's ability to seek out and settle in advantageous habitat, thereby reducing predation and increasing survivorship.