Movements of juvenile Homarus americanus (American Lobster; hereafter lobster) on and around a naturally occurring rock reef were monitored over a 3-year period. Lobsters were sampled with baited traps deployed at each often sites. Catch-per-unit-effort (CPUE) and number of lobsters collected per trap haul was calculated for each sampling event. Physical habitat, visually characterized by underwater video and diver observations, differed among sites. Lobster CPUE was significantly greater at rocky sites (>70% density of cobble and/or boulder) containing complex structure, vertical relief from the seafloor, and colonies of macroalgae, sponge, and hydroids. Lobster CPUE was highest from late June to mid-July. Lobsters ranged from 18 to 82 mm carapace length (CL), with 90.7% of tagged lobsters measuring between 30 to 60 mm CL. Relative lobster abundance remained similar over the course of the study. Catch data were kriged to illustrate spatial patterns of distribution. Over the study period, a total of 934 lobsters were tagged and 66 were recaptured, for an overall recapture rate of 7.1%. The majority of recaptured animals (88%) were found at the original tagging site or adjacent sites, with one lobster remaining at liberty for 397 days. Most juvenile lobsters showed fidelity to their initial site of capture on a small, relatively isolated patch of rock-reef habitat in the central basin of Long Island Sound.