Young injecting drug users (IDUs) are at high risk for a number of negative health outcomes such as hepatitis B, hepatitis C, and human immunodeficiency virus (HIV) infection. However, very little is known about injecting drug-use patterns among this population, particularly with respect to cessation of injection. We sought to identify the factors associated with cessation of injection in a population of young street-based IDUs. A prospective cohort study design was used to assess long-term (> or = 1 year) cessation of drug injection. Data was collected between January 1995 and September 2000 in Montreal, Québec, Canada. Subjects were originally recruited from various street-based outreach programs in Montreal and, for this study, had to have reported injecting drugs within the prior 6 months at baseline or during follow-up and had to have completed at least two semiannual follow-up questionnaires. Cessation incidence rates stratified by duration of injection and adjusted hazard ratios (AdjHRs) were calculated. A Cox proportional hazards regression model was used to identify risk factors independently associated with cessation of drug injection. Of 502 young IDUs, 305 subjects met the inclusion criteria. Cessation of injection for approximately 1 year or more occurred in 119 (39%) of the young IDUs. The incidence of cessation was 32.6/100 person-years but consistently declined as duration of time spent injecting increased. Independent predictors of cessation of injection were currently injecting on a less than monthly or less than weekly basis (HR = 6.4; 95% confidence interval (CI): 3.0-13.6 and HR = 2.4; 95% CI = 1.1-5.3, respectively); currently injecting two or fewer different types of drug (HR = 2.1; 95% CI = 1.1-4.0); currently employed (HR = 1.7; 95% CI = 1.1-2.7); and having at least one parent born outside of Canada (HR = 1.4; 95% CI = 1.1-1.7). Independent predictors of not ceasing injection were currently attending a needle-exchange program (HR = 0.5; 95% CI = 0.3-0.8); and current homelessness (HR = 0.6; 95% CI = 0.4-1.0). The early sharp decline in cessation of drug injection followed by a consistent decrease in this rate suggest difficulties in breaking the habit later on in the drug injecting career. Intensity of drug use and factors which may help to stabilize the social environment of the young IDU may also influence the ability to stop injecting.