Background. Risk perception is a key component of theories of health behaviour upon which knowledge-based prevention strategies for HCV are predicated. Objectives. Determine the relationship between empirical and perceived HCV risk in a Canadian population of illicit opioid users, and differentiate between those who more accurately perceive and those who overestimate or underestimate their risk. Design. Analysis of the OPICAN cohort, a Canadian five-centered study of opioid users not in treatment. Methods. Empirical risk is correlated with perceived risk using Kendall's Tau beta. A multinomial logistic regression of concordance group on age, sex, relationship status, housing and income is performed. Results. Perceived risk is positively associated with empirical risk (Kendall's Tau = 0.085, p = 0.081). Age, sex, housing and relationship are significant predictors of concordance group membership (n = 288). Conclusion. The composition of concordance groups is described, thus providing targets for public health practitioners to further study, and, ultimately, implement tailored HCV prevention interventions.