Mechanical predator removal programs have gained popularity in the United States and have benefited the recovery of several native trout and spring fish. These successes have been limited to headwater streams and small, isolated ponds or springs. Nevertheless, these same approaches are being applied to large river systems on the belief that any degree of predator removal will somehow benefit natives. This attitude is prevalent in the Colorado River mainstem where recovery and conservation programs are struggling to reverse the decline of four endangered fish species. Predator removal and prevention are major thrusts of that work but unfortunately, after 10 years and the removal of >1.5 million predators, we have yet to see a positive response from the native fish community. This leads to the obvious question: is mechanical removal or control in large (>100 cfs base flow) western streams technically or politically feasible? If not, recovery for some mainstem fishes may not be practical in the conventional sense, but require innovative management strategies to prevent their extirpation or possible extinction. This article examines (1) what has been attempted, (2) what has worked, and (3) what has not worked in the Colorado River mainstem and provides recommendations for future efforts in this critical management area.