In a double-blind randomized placebo-controlled trial, 30 cocaine-dependent injecting drug users were assigned to receive 60 mg/day dexamphetamine or placebo for 14 weeks. Immunoassay and mass spectrometric techniques were used to identify cocaine metabolites in urine. Subjects were screened using the Composite International Diagnostic Interview and DSM-IV. To collect pre- and post-self-report data, the Opiate Treatment Index, Brief Symptom Inventory, Severity of Dependence Scale, and visual analogue craving scales were used. The results show that the response to the intervention among a difficult to engage and treat population was encouraging. Outcomes favored the treatment group in terms of cocaine use measured by urinalysis and self-report, crime, cocaine craving, and severity of cocaine dependence, with no improvements observed in the placebo group. The results supported the specific efficacy of dexamphetamine in attenuating cocaine craving. The findings in the psychological dimension of dependence were supported further by declines found in the severity of cocaine dependence scores. The intervention and study design appeared to be acceptable to cocaine users, as evidenced by satisfactory recruitment and retention in an otherwise highly chaotic group. Further controlled studies are advised with adequate power to evaluate the utility of dexamphetamine in the management of cocaine dependence. 1 figure, 1 table, 14 references