The availability of affordable, direct-to-consumer genetic tests has mushroomed, leaving regulation lagging behind. Dozens of companies now offer inexpensive home kits that allow people to spit into tubes, send the samples for DNA analysis and receive a report that allegedly details their ancestry or their possible susceptibility to a long list of disorders that have been linked — often tenuously — to particular genes. But the value of these tests remains debatable, which is why the industry needs a strong set of quality standards and codes of conduct to protect both its consumers and its own credibility.

The UK Human Genetics Commission (HGC) took a welcome step in that direction last month when it issued a set of principles to help guide consumers and to promote high standards and consistency among personal-genomics providers. But the HGC's guiding principles — which are under public review until early December — focus largely on reining in companies' promotional messages so that they reflect the limited utility of genetic testing, and to make would-be customers more aware of what they can realistically expect to learn from the tests. Most DNA testing companies say they are already doing just that, emphasizing that what they provide is information, not medical diagnoses.

The question is what happens if or when prices drop further and the tests become more popular. They are already being marketed over the Internet with little oversight, and it seems likely that increasing numbers of people will be turning to personal-genomics companies in search of definitive answers about how to improve or safeguard their health. But the available answers are rapidly becoming less definitive: the ever-increasing number of genome-wide association studies, which provide a major portion of the genetic markers for disease risks, offer results that even researchers can find hard to interpret, and often flag up markers that are not the most useful predictors of complex traits (see page 712).

This means that customers will frequently receive results telling them only that they face the ambiguous possibility of a somewhat elevated risk of a little-understood disorder. Presumably, most customers in that position will simply be more vigilant about disease screening. But if the ambiguous, slightly elevated risk relates to a frightening condition such as breast cancer, some individuals might feel compelled to undertake drastic and perhaps needless measures, such as prophylactic mastectomy.

The HGC's principles, if adopted, should help to minimize such panic reactions. For example, the HGC advocates that genetic counselling be provided both before and after testing for serious hereditary diseases. But there is room to go further and, on page 724, Craig Venter and his colleagues offer nine recommendations for how to do so. For example, Venter's team urges companies to agree on a core set of non-ambiguous genetic markers — ones that put a carrier at high risk of developing a specific condition. Companies would be well-advised to follow this recommendation, as such an agreement would help to avoid conflicting messages and inconsistent results across the industry.

Ultimately, however, government regulators may feel compelled to step in on the grounds that industry self-policing no longer offers consumers adequate protection. The US Food and Drug Administration has already recruited bioethicist Alta Charo, in part to advise commissioner Peggy Hamburg on a comprehensive approach to regulating these tests.

Government regulators should proceed with care, given the dizzying speed at which the science of personalized genomics is advancing. But in the interim, DNA-test providers should up their game by providing only clinically useful information and spelling out exactly how much biology remains unknown.