We cannot agree with Zheng Wan's view that it is getting harder for scientists in China to access high-quality public data (Nature 520, 587; 2015). There are still technical issues to be resolved, but the trend is towards a greater amount of open data — not less.

The Chinese government has established several national data infrastructures over the past few years that are publicly accessible. These include the National Science and Technology Infrastructures organization (www.escience.gov.cn), which has been sharing data since 2009 in fields such as Earth systems, marine science, meteorology, agriculture, forestry, medicine and health. One of its 23 projects is the National Specimen Information Infrastructure (www.nsii.org.cn), launched in 2013, which currently archives data from more than 10.5 million biological, mineral, rock and fossil specimens.

The degree and success of data sharing also rely on individual scientists' attitudes and practices. Many researchers — including some who are publicly funded in China — unfortunately often resist sharing data, actively or passively. They should instead take more responsibility for cultivating a sustainable, 'bottom-up' culture of sharing.