This study sought to develop a baiting strategy to deliver an oral rabies vaccine to free-ranging coyotes (Canis latrans) in southern Texas. To determine bait longevity, dog foodlard baits were placed (n = 50) on- and off-roads during July 1994 and January 1995. Coyote visitation and uptake rates did not differ between on-road and off-road placement of baits. To evaluate bait stations as possible visual cues, baits were placed out both with (n = 50) and without (n = 50) bait stations. A visual cue of a bait station did not affect coyote response to baits. Bait longevity was shorter during July (< or = 4 days) than January because of consumption of baits by imported fire ants (Solenopsis invicta). The effect of two different bait densities on coyote acceptance rates was determined on six 93.5 km2 study areas in southern Texas. Three study areas received a bait density of 19 baits/km2 (50 baits/mi2) and the remaining three study areas received 58 baits/km2 (150 baits/mi2). Coyote bait uptake rates, based upon the proportion of coyotes marked with either tetracycline hydrochloride or rhodamine B or both, were 83% (n = 99 coyotes) and 87% (n = 101 coyotes) for the 19 and 58 baits/km2 densities, respectively. Bait uptake rates did not differ (P > 0.54) between the two bait densities. Rodents and rabbits, which were fed baits containing tetracycline hydrochloride and a simulated oral rabies vaccine sachet containing rhodamine B, did consume the bait but not the rhodamine B sachet. These animals then were killed and fed to captive coyotes (n = 9). Canine teeth were extracted from coyotes and processed for tetracycline determination. Each coyote tested negative for tetracycline. Therefore, it was unlikely that coyote bait consumption rates were overestimated because of coyotes secondarily marking themselves by ingesting prey items that consumed baits.