Control of big sagebrush (Artemisia tridentata) on heavily infested areas has long been recognized as an effective range improvement practice. By killing this one undesirable species the grazing capacity of rangelands can often be increased manyfold. During the past 20 years several methods have been developed to reduce sagebrush numbers and bring about such range improvement (Pechanec et al., 1954). Generally these methods can be divided into two main categories: those that destroy all existing vegetation, and selective methods that destroy sagebrush without complete destruction of the herbaceous understory. When there are few plants of desirable species, consideration in the selection of a control method generally need be given only to effectiveness of sagebrush removal, relative cost, and development of a satisfactory seedbed. Establishment of forage species will be dependent upon artificial seeding. On the other hand, if a good stand of desirable species is present, it is upon these that increased production should depend; selection of the eradication method should not only consider degree of sagebrush kill and cost but also the effect of treatment upon these associated plants. It is with this question of the effect of treatment on assocJiated fbrage species that this paper is concerned. Considerable information has already been obtained on the effect of various selective methods of sagebrush control. Blaisdell (1953) reported in detail the effect of sagebrush burning on residual vegetation on the Upper Snake River Plains of Idaho. Bohmont (1954), Hyder and Sneva (1956), and Blaisdell and Mueggler (1956) are among those who have studied the effect of the sagebrush-killing 2,4-D sprays on associated vegetation. Burning, spraying, rotobeating, and railing as well as other methods were described by Pechanec et al. (1954), but only general information was presented on the relative effects of these methods on associated vegetation. Evidently little information is available on direct comparisions of various methods of sagebrush removal under carefully controlled conditions. A comparison of four methods of sagebrush control that maintained the native herbaceous understory-burning, rotobeating, railing, and spraying with 2, 4-D-was started in 1952 on the Upper Snake River Plains in southeastern Idaho. The study area is in a fairly homogeneous, dense stand of big sagebrush between 2 and 3 feet in height, on the spring-fall range of the U. S. Sheep Experiment Station near Dubois, Idaho. The topography has only slight relief with little surface drainage; the sandyloam soil is underlain by basaltic lava. Precipiation averages about 13 inches annually. A good understory of native perennial species was present throughout. This herbaceous understory consisted predominantly of bluebunch wheatgrass (Agropyron spicatum), thickspike wheatgrass (A. dasystachyum), Idaho fescue (Festuca idahoensis), plains reedgrass (Calamagrostis montanensis), subalpine needlegrass (Stipa columbiana), blue grasses (Poa spp.), threadleaf sedge (Carex filifolia), fleabane (Erigeron corymbosus), lupines (Lupinus spp.), and minor amounts of other grasses and forbs. Besides the dominant big sagebrush, the shrubby vegetation included spineless gray horsebrush (Tetradymia canescens), downy rabbitbrush (Chrysothamnus puberulus), and antelope bitterbrush (Purshia tridentata). A more detailed listing of species and relative amounts can be found in Table 2.