Several investigators have described an interaction between muscle relaxants and corticosteroids which have showed different results. The exact mechanism of this section is not clear and there conflicting results have further confusion. For the confirmation on how to interact these two drugs in patients without influencing factors to the neuromuscular transmission, corticosteroids(hydrocortisone, methyl prednisolone) 0.5mg/kg and 5.0 mg/kg mixed in 20ml of normal saline with micro-dose of various muscle relaxants(vecuronium 0.1mg/kg/30, d-Tubocurarine 0.4mg/kg30, succinylcholine 1mg/kg/30) were administered respectively into the vein at the distal portion of a pneumatic tourniquet applied on the upper arm with higher pressure than arterial pressure. This was to study the dose response according to recovery time between the experimental troup and the control group. Neuromuscular block was measured by recording the twitch response following ulnar nerve stimulation by EMG(ABM, Datex Co., 2Hz 30 mA supermaximal voltage). And at over 95% depression of the twitch height, following administration of muscle relaxants alone(control group) or muscle relaxants combined with corticosteroids(study group), the tourniquet was released, and recovery time from 25% to 75% recovery of twitch height was measured according to recovery index(RI). The results obtained were as follows: 1) In the group which 0.5mg/kg of hydrocortisone was administered. the RI of vecuronium and d-tubocurarine was shorten to 3.47±1.43 and 9.30±1.48 minutes, and in hydrocortisone 5.0mg/kg, was prolonged to 12.30±2.18 and 17.96±0.62 minutes, compared with control group(8.08±1.48 in vecuronium and 13.43±2.32 in d-Tubocurarine). 2) Methyl prednisolone also tended to short the RI in 0.5 mg/kg of corticosteroids and to prolong in 5.0 mg/kg, but not significant. 3) Corticosteroids not to depend on dose prolonged the RI from succinylcholine induced neuromuscular blocks. 4) Plasma cholinesterase values following injection of hydrocortisone 0.5mg/kg and 7.0mg/kg were not changed significantly. 5) Hydrocortisone alone had significant effect on twitch tension, increased in dose of 0.5mg/kg and depressed in dose of 5.0mg/kg. Conclusively, in the interaction between corticosteroids and muscle relaxants, a combined effect of antagonism in small dose and enhancement in large dose of corticosteroids to non-depolarizing neuromuscular blocking action and a potenciated effect to depolarizing neuromuscular blocking action, seems to be possible. We recommend careful monitoring of neuromuscular transmission in patients receiving corticosteroids when neuromuscular blocking agents in used for muscle relaxation.