Food-searching behavior of two adult male Ferruginous Hawks (Buteo regalis) was studied by direct observation in Curlew Valley, Utah-Idaho, during the nesting seasons of 1974 and 1975. This paper analyzes the effects of vegetation type, prey distribution, cover density, previous experience, and time and energy budgets upon the predators' choice of hunting methods and hunting sites. Both hawks hunted actively throughout the day (0600 to 2100 hours MDT). Male 1 (1974) and Male 2 (1975) captured an average of 8.7 ± 2.0 and 9.0 ± 1.5 prey items per day, respectively, during the 4-wk observation period in each year. The hawks hunted by four distinguishable methods: (1) from a perch, (2) from the ground, (3) from low-altitude (active) flight, and (4) from high-altitude (soaring) flight. The birds made a total of 808 observed strikes and were successful in securing prey in 129 (16.6 percent) of those strikes. Success rate (successful strikes/total attempted strikes) varied significantly with hunting method. The hawks were selective in their use of vegetation types for hunting. Both birds used the "bare ground" and "pasture" types more than expected by chance. Male 2 also over-used the "alfalfa" type. Both hawks significantly under-used the "grass-shrub," "grass or grain," "old field," "juniper," and "rush-grass" cover types. The differences in use of vegetation types could not be related to differences in prey biomass. The birds were selective in their use of areas which differed in density of vegetative cover. Areas with no cover were used by both hawks more than expected by chance; all other areas were under-used with the single exception of "dense" cover in 1975. Previous experience was important to the birds' choice of hunting sites and hunting methods. Each hawk returned directly to the site of its last prey capture on more than half of its hunting forays. Similarly, each hawk tended to initiate a new hunting effort by using the hunting method that had been successful on its previous hunt. In 1975, Male 2 switched its emphasis to a new major hunting area during the fourth week of observation. The apparent reason for the switch was a decline in success at the first hunting area due to the growth of vegetation which concealed prey. Males 1 and 2 averaged 0.94 ± 0.43 and 1.27 ± 0.55 prey captures per hour of hunting time, respectively. Capture rates (captures/time) varied with hunting method. The amount of time a hawk spent hunting by each method was not related to its capture rate by that method. However, each bird's total use of sit-and-wait hunting (from a perch or from the ground) and of hunting from low-altitude flight was proportional to estimates of the number of captures per unit of energetic cost for those methods. Both hawks hunted from high-altitude flight more than expected on the basis of their catch/cost ratios for that technique. This suggests that soaring was not exclusively a hunting technique, but that it had some additional purpose unrelated to food gathering. There was some evidence that the hawks used major hunting sites which provided high benefit/cost ratios relative to the rest of their foraging ranges. Both hawks foraged efficiently by concentrating more of their foraging time on the hunting method and in the hunting sites which yielded the highest number of prey captures per unit of energy ex-pended. The birds' use of secondary hunting areas may have contributed to their long-term foraging efficiency by informing them of changes in the foraging potential of different portions of their home ranges. Such flexible foraging patterns may be critical to the survival of predators which rely on highly variable small-mammal populations.