Can J Psychiatry 2008;53(12):793-794 If cannabis is a contributory cause of psychosis does it follow that we should recriminalize its use in countries that have removed criminal penalties? Leading politicians in Australia and the United Kingdom think we should. We disagree. We accept it is more likely than not that cannabis is a contributory cause of psychosis but we doubt that recriminalizing its use will achieve the goal of reducing cannabis use or cannabis-related psychosis at an acceptable social and economic cost. We accept that regular, especially daily, cannabis use probably is a contributory cause of psychosis for reasons spelt out in detail elsewhere.1 Briefly, these are: * Regular cannabis use and psychotic symptoms and disorders are associated in population surveys and in clinical settings. * Longitudinal studies of representative samples of young people have consistently found that cannabis use at baseline predicts an increased risk of psychotic symptoms or disorders. * This association has persisted after controlling for plausible confounding variables such as personality traits and other drug use.2 * A causal relation is biologically plausible given evidence of interactions between the cannabinoid and dopaminergic neurotransmitter systems.1 Some leading researchers remain skeptical about this evidence. They argue that the modest size of the relation (relative risk, 2) and its attenuation after adjustment for confounders suggests that residual confounding provides the best explanation of the association.3 For the purposes of this article, we will assume that the relation is causal, in the sense that cannabis is a contributory cause that acts in concert with preexisting vulnerability and other unknown factors to increase the risk of psychotic symptoms. Under current patterns of cannabis use (and assuming the relation is causal), the attributable risk of psychosis from cannabis use is around 10%. Epidemiologic modelling indicates that any effects of a causal relation of this size on the incidence or prevalence of psychosis may be difficult to detect.4 Cannabis use among young people is unlikely to produce an epidemic of schizophrenia, as is suggested in media stories, but any increase in such disorders would nonetheless be of public health and humanitarian concern because of the adverse effects that these disorders have on the life chances of the young people affected by them. Concern about any psychotogenic effects of cannabis has been heightened by reported increases in the tetrahydrocannabinol (THC) content of cannabis products that are nowadays available to young people. Similar claims have been made for over 20 years with a lack of supportive evidence, but there is now evidence that THC content has increased in recent decades from 3% to 6% in the United States.5 Such an increase reflects increased efficiency in the cultivation of high THC yield cultivars and hydroponic methods of cultivating sinsimella plants. It may also reflect increased demand for higher THC content cannabis products from regular users and increased regular use of cannabis beginning at an earlier age.5 It should be noted that, to the extent cannabis users titrate their doses, the result of higher THC cannabis may be a reduction in some adverse physical health effects for regular users. While accepting the likelihood that cannabis makes a modest contribution to psychosis risk, we nonetheless argue that policy-makers should avoid an unreflective leap to the policy conclusion that recriminalizing cannabis use is the obvious remedy. We do so for 5 main reasons. First, there is no evidence that removing criminal penalties has had any effects on rates of cannabis use. In the mid-1970s in the United States, rates of cannabis use increased at the same rate in states that legislated to remove criminal penalties as it did in those that retained them.6 The same was true in the Netherlands in the mid- 1 970s. …