A DECADE ago, Lovelock and Whitfield1 raised the question of how much longer the biosphere can survive on Earth. They pointed out that, despite the current fossil-fuel induced increase in the atmospheric CO2 concentration, the long-term trend should be in the opposite direction: as increased solar luminosity warms the Earth, silicate rocks should weather more readily, causing atmospheric CO2 to decrease. In their model1, atmospheric CO2 falls below the critical level for C3 photosynthesis, 150 parts per million (p.p.m.), in only 100 Myr, and this is assumed to mark the demise of the biosphere as a whole. Here, we re-examine this problem using a more elaborate model that includes a more accurate treatment of the greenhouse effect of CO2 (refs 2-4), a biologically mediated weathering parameterization, and the realization that C4 photosynthesis can persist to much lower concentrations of atmospheric CO2(<10 p.p.m.)5,6. We find that a C4-plant-based biosphere could survive for at least another 0.9 Gyr to 1.5 Gyr after the present time, depending respectively on whether CO2 or temperature is the limiting factor. Within an additional 1 Gyr, Earth may lose its water to space, thereby following the path of its sister planet, Venus.