American woodcock (Scolopax minor) have experienced long-term population declines due to habitat loss. While significant research has occurred on breeding and wintering grounds, little is known about spring migratory ecology. This study assessed nocturnal roosting habitat of American woodcock through the use of night spotlighting techniques. The study was conducted on the Blue Grass Army Depot and the Central Kentucky Wildlife Management Area, both located in Madison County, Kentucky. A total of 84 field, 421 ha, of four habitat types, burned, grazed, hayed, old/fallow, were searched for woodcock. Roost sites were marked and the vegetation of each site was compared between age classes, sexes, and between occupied and random unoccupied locations. The following vegetation parameters were assessed: percent cover (bare soil, grass, forbs/gramanoids, shrub/sapling/vine, and litter), litter depth, dominant plant height, dominant plant species, distance to escape cover, and percent vertical cover. There was no significant difference (P > 0.05) between sexes or age classes for any of the habitat variables assessed. Logistic regression analysis indicated the best predicators of whether a woodcock would be present at a roost site were percent bare, grass, graminoids, and woody vegetation, litter depth, dominant height of vegetation, distance to escape cover, visual obstruction of escape cover from 0-20cm, and distance to field edge. To predict density of woodcocks in each field I used step-wise regression analysis, which indicated the best model for predicting woodcock density per field incorporated percent litter at roost site, litter depth, distance to escape cover, visual obstruction of escape cover from 0-20 cm, and visual obstruction of escape cover from 50-100cm. Woodcock selected fields and roost sites with varying heights of vegetation that satisfied ecological needs during spring migration, i.e. resting, loafing, breeding, foraging, predator and weather avoidance.