Use of sheep to control weeds in a ryegrass-white clover pasture in which Knob Cone-Monterrey hybrid pine trees had been planted was investigated in Corvallis, Oregon during 1997-1998. Treatments consisted of an ungrazed control (C) and grazing applications to achieve 25 (L), 50 (M) or 75 (H) percent utilization of the understory vegetation. The treatments were applied three times between May and July of 1998 (T1, T2, and T3). Cover and phenological stage by plant group was measured after each treatment application. Two species of interest, Himalayan blackberry and bull thistle, were also studied independently of the other forbs. Their growth was measured at T1, T2, T3 and also in October of 1998 (T4). All yearling ewes were weighed at the beginning of the trial, before T1, and again after T3. Sheep used in the trial were also weighed after T2. Initially understory vegetation covered about 90% of the ground, approximately 20% of which was perennial ryegrass (RYE). Incidence of other perennial grasses (OPG) remained constant at about 20% where ungrazed but declined from 20% to 10% during the study in all grazed treatments. Annual grasses (AG) performed similarly to OPG, remaining at 25% in C throughout the trial but declining to 8% or less in all grazed treatments by T3. Forb cover was variable throughout the site, ranging from about 20 to 50% before initial grazing. Grazing at T1 reduced forb cover by 44% (L) to 80% (H) from initial levels. Forbs did not fully recover by T2 and constituted about 25% of the cover in C and L and 10% in M and H. Grazing at T2 reduced forb presence by 25% in L, with little change in M and H due to the low presence of forbs in those treatments. Late season forbs began to appear by T2 and were dominant by T3 when grazing reduced total forb cover to 15% in L and