90% of the world's rice is grown and consumed in Asia, with each hectare of rice-producing land providing food for 27 people. By 2050, because of population growth and increasing urbanisation, each remaining hectare will have to feed at least 43 people. This means that yields must be increased by at least 50% over the next 40 years to prevent mass malnutrition for the 700 million Asians that currently rely on rice for more than 60% of their daily calorific intake. Since predictive models suggest that yield increases of this magnitude can only be achieved by improving photosynthesis, and because evolution has increased photosynthetic efficiency by 50% in the form of the C4 pathway, one solution is to generate C4 rice. However, this is an ambitious goal that requires proof of concept before any major investment of time and money. Here, we discuss approaches that should allow proof of concept to be tested.