Two experiments examined the effects of methylphenidate in male and female patients enrolled in an outpatient treatment program for primary cocaine dependence. The first study was a component of a double-blind efficacy trial wherein 57 patients were first tested in a human laboratory for their initial responsiveness to medication. Patients were randomly assigned to receive either placebo or methylphenidate treatment and received their first dose in the human laboratory environment before continuing in outpatient treatment. Methylphenidate was given as a 20-mg sustained-release dose (twice daily) plus an additional 5-mg immediate-release dose combined with the morning dose. Methylphenidate increased heart rate and subjective ratings; however, the subjective effects were primarily of a 'dysphoric' nature, and significant effects were limited to increases in anxiety, depression, and anger on the Profile of Mood States; shaky/jittery ratings on a visual analog scale; and dysphoria on the lysergic acid diethylamide (LSD) scale of the Addiction Research Center Inventory. Methylphenidate did not increase cocaine craving nor ratings suggesting abuse potential (i.e., Morphine-Benzedrine Group or drug-liking scores, etc.). None of the drug effects observed in the human laboratory was of clinical concern, and no subject was precluded from continuing in the outpatient study. After outpatient treatment completion, 12 patients were brought back into a second double-blind human laboratory study in which three doses (15, 30, and 60 mg) of immediate-release methylphenidate were administered in an ascending series preceded and followed by placebo. Methylphenidate produced dose-related increases in heart rate, subjective ratings of shaky/jittery, and LSD/dysphoria without significantly altering cocaine craving or stimulant euphoria ratings. These results suggest that stimulant substitution-type approaches to the treatment of cocaine dependence are not necessarily contraindicated because of cardiovascular toxicity or medication abuse potential. However, they also suggest that the subjective effects of methylphenidate may not be positive enough for an adequate replacement approach.