Intravenous cocaine (COC) and morphine (MOR) combinations ("speedballs") were administered to nine research volunteers during 12 daily experimental sessions. During each session, a single dose of MOR sulfate (0, 5 and 10 mg/70 kg) was administered i.v. in combination with a single dose of COC hydrochloride (0, 8, 16 and 32 mg/70 kg). All possible COC-MOR combinations were each tested once. Measures of subjective effects were obtained repeatedly, heart rate, diastolic and systolic blood pressure were sampled every 2 min and blood levels of COC and MOR were obtained at three time points. Both COC and MOR alone significantly increased peak heart rate, diastolic pressure and systolic pressure. The cardiovascular effects of COC were larger and longer in duration than morphine, as indicated by larger area under the curve measures compared to MOR. COC alone increased visual analog scales ratings of "stimulated" and A scores of the Addiction Research Center Inventory, whereas MOR alone increased opiate symptoms and ratings of "sedated" and both drugs increased ratings of "high." For the most part, the cardiovascular effects of COC-MOR combinations could be predicted by the effects of COC alone, and the subjective effects of the combinations could be predicted by either drug alone. The drug combination produced increases in cardiovascular and subjective effects that were less than predicted by an effect addition model of drug interaction. The disposition of both COC and MOR was unaffected by the presence of the other drug. These results suggest that the self-administration of this drug combination may be related to a unique profile of opiate and stimulant effects instead of an alteration in the effects produced by either drug alone.