Agonists, or "replacement medications," are useful adjuncts in treatment of opiate and nicotine dependence. They have not been systematically examined in cocaine dependence. Results of early open trials with methylphenidate for treatment of cocaine dependence were equivocal. Twenty-four cocaine-dependent subjects were enrolled in an 11-week double-blind, placebo-controlled study of methylphenidate. Assignment was random. Intake included a 2-day human laboratory procedure in which subjects received initial doses of methylphenidate or placebo. Subjects attended the clinic Monday through Friday and received oral doses of methylphenidate (5 mg plus 20-mg sustained release) or placebo at 8:00 a.m., with afternoon and weekend take-home doses (20 mg sustained-release or placebo) provided in Medication Events Monitoring System bottles to monitor compliance. Clinic visits included therapy sessions, electrocardiograms, self-report measures, and twice-weekly urine screens. The two groups were equivalent in terms of retention (methylphenidate 48% and placebo 42%) and had similar cocaine use outcomes (40% benzoylecgonine-positive urine screens). There were no significant adverse effects. The doses were sufficient to permit detection of psychoactive effects ("stimulant," "more energy") and side effects ("jitteriness," "eating less") without increased "craving." Additional medications with different effects profiles are being studied to further evaluate the replacement model in cocaine dependence. (J Clin Psychopharmacol 1997;17:485-488).