The effects of either clonidine or epinephrine into local anesthetics administered into brachial plexus sheath were evaluated in 42 patients who underwent surgery of the upper limb. All patient received 0.5 ml/kg of 2:1 mixture of bupivacaine and lidocaine injected into the brachial plexus sheath, using the subclavian perivascular technique. The patients were randomly allocated into two groups; Group I(n=25) received $150{mu}g$ of clonidine hydrochloride, and Group II(n=27) received $200{mu}g$ of epinephrine. The duration of analgesia and the degree of sedation reflecting the systemic effect of clonidine were assessed. The block produced by the addition of clonidine was longer($100.3{pm}469.8$ vs $648.8{pm}192.1$ min) and superior to that by epinephrine(p < 0.05). The highest degree of sedation was achieved about 20 minutes after block, which roughly equals the time required for intramuscular clonidine to show the similar effect. The author concludes that the injection of clonidine mixed to local anesthetics into the brachial plexus sheath prolongs analgesia than that of epinephrine, but this prolongation may be due to the systemic effect of clonidine.