Background-There is a growing concern surrounding crystal methamphetamine use in Canada despite surprisingly little empirical data to support such claims. We evaluated the trends in crystal methamphetamine injection and factors associated with injection of the drug among a cohort of injection drug users (IDU) in Vancouver.Methods-We conducted a prospective analysis of factors associated with crystal methamphetamine injection among participants enrolled in the Vancouver Injection Drug Users Study (VIDUS). Since serial measures for each individual were available, variables potentially associated with crystal methamphetamine injection were evaluated using generalized estimating equations (GEE) with logit link for binary outcomes.Results-Overall, 1587 IDU were enrolled into the VIDUS cohort between May 1996 and December 2004. The proportion of IDU who reported injecting crystal methamphetamine during the last six months increased during the study period (Cochran-Armitage trend test, p < 0.001). In multivariate GEE analyses, crystal methamphetamine injection was independently associated with younger age (Adjusted Odds Ratio [AOR]: 4.77, 95% Confidence Interval [CI] = 3.40 -6.70), Caucasian ethnicity (AOR = 2.21, 95% CI = 1.57 -3.12), syringe borrowing (AOR = 1.62, 95% CI = 1.22 -2.13), and syringe lending (AOR = 1.40, 95% CI = 1.02 -1.86).Interpretation-There was a significant trend towards increasing crystal methamphetamine injection in this setting and elevated HIV risk behavior and younger age were independently associated with crystal methamphetamine injection. Given that banning precursor chemicals has had a limited effect on reducing methamphetamine supply in other jurisdictions, pragmatic and effective interventions are needed to address the growing use of this drug.