The design of more than 90% of 2235 randomised controlled trials published in Chinese medical journals was flawed, concludes a review (Trials 2009;10:46, doi:10.1186/1745-6215-10-46). Researchers trawled a Chinese national database for studies of 20 common diseases published between 1994 and 2005. Only 207 of the studies used accepted randomisation methods. Data from falsely reported trials can mislead healthcare providers, consumers and policy makers. In a recent Lancet article, Jia He and colleagues at the Second Military Medical University in Shanghai, said, “Over the past 20 years biomedical articles authored and published by Chinese researchers have improved greatly in quality” (2009;373:2091-3). (BMJ 2009;339:b2729)