Cannabis use may be a risk factor for schizophrenia. Part of this association may be explained by genotype–environment interaction, and part of it by genotype–environment correlation. The latter issue has not been explored. We investigated whether cannabis use is associated with schizophrenia, and whether gene–environment correlation contributes to this association, by examining the prevalence of cannabis use in groups with different levels of genetic predisposition for schizophrenia.

Case-control study of first-episode schizophrenia. Cases included all non-Western immigrants who made first contact with a physician for schizophrenia in The Hague, The Netherlands, between October 2000 and July 2005 (n=100; highest genetic predisposition). Two matched control groups were recruited, one among siblings of the cases (n=63; intermediate genetic predisposition) and one among immigrants who made contact with non-psychiatric secondary health-care services (n=100; lowest genetic predisposition). Conditional logistic regression analyses were used to predict schizophrenia as a function of cannabis use, and cannabis use as a function of genetic predisposition for schizophrenia.

Cases had used cannabis significantly more often than their siblings and general hospital controls (59, 21 and 21% respectively). Cannabis use predicted schizophrenia [adjusted odds ratio (OR) cases compared to general hospital controls 7.8, 95% confidence interval (CI) 2.7–22.6; adjusted OR cases compared to siblings 15.9 (95% CI 1.5–167.1)], but genetic predisposition for schizophrenia did not predict cannabis use [adjusted OR intermediate predisposition compared to lowest predisposition 1.2 (95% CI 0.4–3.8)].

Cannabis use was associated with schizophrenia but there was no evidence for genotype–environment correlation.