Arterial plasma drug concentrations should be a strong predictor of resultant physiological and behavioral effects of smoked drugs. Nine healthy male volunteers, who were current users of smoked and i.v. cocaine, participated in a study directly comparing arterial and venous cocaine plasma concentrations after smoked and i.v. cocaine. Each volunteer was first tested under all dosing conditions without an arterial catheter (phase I), to determine whether subjects could tolerate each cocaine dose, before testing with an arterial catheter (phase II). Phase II consisted of two test days, each consisting of either four smoked-cocaine sessions (sham, 12.5, 25 and 50 mg) or four i.v. cocaine sessions (0, 8, 16 and 32 mg) in ascending order, spaced 90 min apart. For the two highest doses for each route, arterial and venous blood samples were taken simultaneously before drug administration, during drug administration and frequently after drug administration. At the same time, cardiovascular effects and subjective effects were measured. Arterial cocaine concentrations were substantially higher than venous cocaine concentrations after both routes of cocaine administration. After either smoked or i.v. cocaine, maximal arterial cocaine concentrations occurred within 15 sec, whereas maximal venous cocaine concentrations occurred within 4 min. The onset of cardiovascular and subjective effects was also rapid, and arterial cocaine concentrations tended to account for these effects to a greater extent than did venous cocaine concentrations. Even though arterial and venous cocaine concentrations were lower after smoked cocaine, compared with i.v. cocaine, the magnitudes of cardiovascular and subjective effects were similar, suggesting that smoked cocaine may produce a greater effect at similar concentrations. These findings do not support the hypothesis that the greater effects observed after smoking are a result of smoked cocaine reaching the brain faster than i.v. cocaine as measured by peripheral arterial plasma concentrations.