BACKGROUND: Cocaine dependence is an increasingly prevalent disorder for which no medication is approved yet. Likewise opioid for heroin dependence, replacement therapy with psychostimulant could be efficacious for cocaine dependence. OBJECTIVES: To ascertain the efficacy of psychostimulants for cocaine dependence on cocaine use, sustained cocaine abstinence and retention in treatment. The influence of type of drug, comorbid disorders and clinical trial reporting quality over psychostimulants efficacy has also been studied. SEARCH STRATEGY: MEDLINE, EMBASE, PsycINFO, CENTRAL, references of obtained articles and experts in the field. SELECTION CRITERIA: Randomized parallel group controlled clinical trials comparing the efficacy of a psychostimulant against placebo have been included. DATA COLLECTION AND ANALYSIS: Two authors evaluated and extracted data. The Relative Risk (RR) was used to assess dichotomous outcomes except for adverse event (AE) induced dropouts for which the risk difference (RD) was preferred. The Standardized Mean Difference (SMD) was used to assess continuous outcomes. To determine the influence of moderating variables, a stratified analysis was conducted. Funnel plots were drawn to investigate the possibility of publication bias. MAIN RESULTS: Sixteen studies have been included, which have enrolled 1,345 patients. Seven drugs with psychostimulant effect or metabolized to a psychostimulant have been investigated: bupropion, dexamphetamine, methylphenidate, modafinil, mazindol, methamphetamine and selegiline. Psychostimulants did not reduce cocaine use (SMD 0.11, 95%CI: -0.07 to 0.29), showed a statistical trend over improving sustained cocaine abstinence (RR 1.41, 95%CI: 0.98 to 2.02, p=0.07) and did not improve retention in treatment (RR 0.97, 95%CI: 0.89 to 1.05). The proportion of AE induced dropouts was similar for psychostimulants and placebo (RD 0.01, 95%CI: -0.02 to 0.03). When the type of drug was included as a moderating variable, it was shown that the proportion of patients achieving sustained cocaine abstinence was higher with bupropion and dextroamphetamine, and also with modafinil, at a statistical trend of significance, than with placebo. Nevertheless, no studied drug was efficacious on any of the remaining outcomes. Besides, psychostimulants appeared to increase the proportion of patients achieving sustained cocaine and heroin abstinence amongst methadone maintained dual heroin-cocaine addicts. The main findings did not seem to be influenced by clinical trial reporting quality. No evidence of publication bias was found. AUTHORS' CONCLUSIONS: This review found mixed results, therefore evidence of the efficacy of psychostimulants for cocaine dependence is inconclusive. Nevertheless promising results exist for methadone maintained dual heroin-cocaine addicts and for some specific drugs such as dexamphetamine and bupropion.