In recent years, societal attitudes to recreational drugs have changed considerably in some parts of the world. Perhaps the best example is cannabis, a widely used substance that has been decriminalised or legalised in various jurisdictions from the US to Portugal. However, even in places where such shifts have taken place, recreational drugs still remain largely taboo, meaning people are likely to be unforthcoming when asked about their own drug-taking habits. This is a problem for health professionals and policy makers trying to establish effective measures to deal with the use of such drugs. While attitudes to these substances are indeed changing, the fact remains that even if used in a regulated context recreational drugs can have all sorts of harmful effects. Making effective interventions requires knowledge of the demographic and individual factors affecting drug use, and to do this researchers need access to accurate data. Misreporting of drug use among survey participants can be a major barrier to the collection of such data, and it is one that Professor Mark Harris, an expert in econometrics at Curtin University in Perth, Australia, has set out to overcome.For Harris, it’s obvious that there can be many reasons underlying someone’s decision to misreport their use of drugs. ‘Clearly participants, especially substance users, are faced with numerous concerns about reporting their use accurately,’ he says, highlighting the perceived social undesirability of such substances, as well as social stigma and potential legal ramifications. ‘Put simply, jointly these mean that the raw numbers obtained from large scale surveys regarding the number of users in the population, and the consumption levels of such (and so on), are likely to be somewhat unreliable.’ Specifically, he says, such surveys are likely to underestimate the prevalence of these behaviours. The article explores some of the research to address the question of how to deal with the ‘excess 0s’ (resulting from false reporting) that arise in survey data owing to these factors.