Transgenic crops producing Bacillus thuringiensis (Bt) toxins for insect control have been successful, but their efficacy is reduced when pests evolve resistance. To delay pest resistance to Bt crops, the U.S. Environmental Protection Agency (EPA) has required refuges of host plants that do not produce Bt toxins to promote survival of susceptible pests. Such refuges are expected to be most effective if the Bt plants deliver a dose of toxin high enough to kill nearly all hybrid progeny produced by matings between resistant and susceptible pests. In 2003, the EPA first registered corn, Zea mays L., producing a Bt toxin (Cry3Bb1) that kills western corn rootworm, Diabrotica virgifera virgifera LeConte, one of the most economically important crop pests in the United States. The EPA requires minimum refuges of 20% for Cry3Bb1 corn and 5% for corn producing two Bt toxins active against corn rootworms. We conclude that the current refuge requirements are not adequate, because Bt corn hybrids active against corn rootworms do not meet the high-dose standard, and western corn rootworm has rapidly evolved resistance to Cry3Bb1 corn in the laboratory, greenhouse, and field. Accordingly, we recommend increasing the minimum refuge for Bt corn targeting corn rootworms to 50% for plants producing one toxin active against these pests and to 20% for plants producing two toxins active against these pests. Increasing the minimum refuge percentage can help to delay pest resistance, encourage integrated pest management, and promote more sustainable crop protection.