Flight range, dispersal activity, and habitat preference were determined for the horse fly,Tabanus abactorPhilip, in north central Oklahoma, using a mark/recapture technique. In 1982 and 1983, engorged and nonengorgedT. abactorwere marked and released at the same release site. Horse flies were recaptured in 12 dry-ice-baited Malaise traps located 0.4 and 0.8 km from the central release site in 1982, and in 20 traps located 0.4–2.4 km from the release site in 1983. A total of 45,153T. abactorwere marked and released, with a mean recapture rate of 6.12%. Recapture rates of ca. 9 and 4% were achieved for the engorged and nonengorged flies, respectively. Engorged flies were recaptured in greatest numbers on days 3 and 4 after release, whereas nonengorged flies were recaptured primarily on days 1 and 2 after release. Dispersal of the engorged flies was the same as that of the nonengorged flies once digestion of the blood meal had occurred. Marked flies were recaptured at all distances from the release site, but over 85% of those recaptured were collected within 8 d after release.