We studied the micro- and macrohabitat preferences of black rats (Rattus rattus) and house mice (Mus musculus) on St. Eustatius, a small, inhabited Caribbean island. Our study builds upon a preliminary assessment of invasive alien rodents on St. Eustatius, which has no extant native rodent species. We deployed tracking tunnels (baited cards with ink left overnight to track animal visitors to the card) in 12 macrohabitats to determine the presence and relative abundance of rodent species. Data were collected between July 2017 and May 2019. We collected data in 5 × 5 meter (n = 120) plots to determine whether grass cover, leaf litter cover, number of living trees, and other microhabitat components influenced habitat use by rats and mice. Rats were more likely to occur in microhabitats with more live trees, but with fewer logs and less grass and bare ground. Mice were more likely to occur in microhabitats with more bare ground. Rat tracks were recorded in all macrohabitats, whereas mouse tracks were detected in all but four, namely: C, M2, M3, and M5. Based on tracking tunnel data, the relative abundance of rats and mice ranged from 0% to 70%, respectively, in all macrohabitats over the study period. Rat relative abundance in forest macrohabitats differed significantly from that in other macrohabitats. Rat relative abundance in regenerating grassland differed significantly from that in other macrohabitats. There is evidence of habitat selection by rats at the micro- and macrohabitat scale, which could be linked to food availability. However, our study revealed only weak or no patterns of habitat selection by mice. Rat relative abundance differed significantly between the second and third assessment, and between the fourth and fifth assessment. Mouse relative abundance differed significantly between the first and second assessment.