Wu and Zhao (2011) argue that pressures and temperatures for mantle xenoliths are best determined by combining Taylor's (1998) garnet-orthopyroxene barometer and two-pyroxene thermometer. Here we show that this choice may lead to grossly erroneous P-T estimates for some peridotitic compositions and increased scatter around xenolith geotherms. We contend that the Nickel and Green (1985) barometer remains a better choice than the Taylor (1998) barometer for thermobarometry of garnet-facies mantle peridotites and pyroxenites, at least for orthopyroxenes with moderate sodium contents. For highly-sodic orthopyroxene compositions, the reliability of the Nickel and Green (1985) barometer and of its modifications proposed by Carswell and Gibb (1987) and Carswell (1991) still awaits detailed evaluation. (C) 2011 Elsevier B.V. All rights reserved.