Scent stations, passive hair-snaggers and howl surveys were evaluated as possible survey methods for monitoring relative abundance of coyotes (Canis latrans) in Great Smoky Mountains National Park (GSMNP) prior to the reintroduction of red wolves (Canis rufus) from January 1990 to April 1991. Scent station nights (N = 198), passive hair-snaggers (N = 70), and howl surveys (N = 197) produced 1 (0.5%), 0 (0.0%), and 35 (17.8%) coyote responses, respectively. Scent stations and hair-snaggers proved ineffective for monitoring coyotes at current population levels. Howl surveys elicited responses from approximately 21 coyotes at 12 locations indicating the feasibility of designing and implementing a standardized survey to monitor the relative abundance of coyotes over time or from area to area. Twenty-seven responses were elicited from coyotes in the Cades Cove section of GSMNP for a coyote index of 22.9% and 8 responses from coyotes outside Cades Cove for a coyote index of 10.1%. Preliminary estimates from 2 indices of relative coyote abundance ranged from 1/13.2 km2 to 1/39.7 km2. Wildlife managers and researchers must accept a wide margin of error if surveys of relative abundance are used for coyotes and other wide ranging carnivores in the southern Appalachians.