State farms, modelled on those successful in northeastern China, could meet burgeoning grain demand in the rest of the country over the next 20 years.

China already grows more than 95% of its own grain, providing food security to 22% of the world's population even though it has only 9% of the world's cultivated farmland (see http://go.nature.com/zhKwtg; http://go.nature.com/fA9TXv). But we estimate that a total of almost 100 billion kilograms more grain will be needed by 2030 (X. Liu and G. Wang Rhizosphere and Yield Physiology of Crops 8–9, Science Press, 2010; in Chinese).

This rising demand will call for a 1% annual increase in total grain production from the same 120 million hectares of cultivated farmland currently available. The only way to step up crop yields is by intensifying production.

State farms in the 'bread basket' of northeastern China have almost doubled their crop yields per unit area compared with the region's local farms. Their average increase in yield during the past 15 years for maize (corn), soya bean and rice, for example, has been 32.6%, 27.8% and 19.9%, respectively. Owing to their greater land area, had local farms adopted similar technology and operational management, their annual grain yield would have been boosted by 82.3%, 66.1% and 32.2%, respectively, for the same crops (Heilongjiang Statistics Yearbook 2008; in Chinese). That is potentially 1.6 times more grain from the same farmland.

Extending this model to other regions in China will bring challenges, including declining water levels, a surplus of farm labour and the need to develop new services to accommodate changes in agricultural structure.

Nonetheless, there is an opportunity here for China to solve its future grain demand through technological solutions and by rationalizing institutional and organizational irregularities (D. W. Bromley Science 328, 169; 2010). Meanwhile, it could reassess its policies for land management, urbanization processes and subsidies.