Allegheny woodrat (neotoma magister) populations in the northern and western limits of the range have been greatly reduced in recent years, increasing the need to locate and monitor both threatened and seemingly stable populations. We tested the feasibility of predicting areas of suitable habitat for the woodrat in the Daniel Boone National Forest (DBNF) by using a Geographic Information System model. Several themes depicting woodrat habitat variables were overlaid to produce a comprehensive map displaying likelihood of woodrat occurrence. Logistic regression analysis was used to determine effect of each habitat variable on woodrat occurrence based on a sample of 394 known woodrat occurrence sites, 511 random sites, and habitat data including slope, landuse, site geology, forest cover, and locations of forest openings, clifflines, streams, and roads. The resulting habitat model correctly classified 97% of the 416 independent woodrat locations at the 0.50 probability level. This habitat model will provide an efficient, cost-effective method for searching out new woodrat locations, monitoring and analyzing previously known locations, managing DBNF to maintain existing habitat, and restoring previous habitat.