It was demonstrated by Bowman et al (1) that an intravenous injection of epinephrine (EPI) caused an increase or a decrease in the tone of the chronically denervated cat mus cles and the effects were mediated by the (l-receptor. In our previous study in vivo (2), denervated rat muscles responded to EPI by a contracture mainly through the a-receptor, while rabbit muscles responded by a fall in the resting tension through both a and fl-re ceptors. The reason for the discrepancy is not clear, though it may in part be due to spe cies difference. In the present paper, effects of EPI and other adrenergic amines were fur ther studied in vitro using chronically denervated rat muscles.The sciatic nerve of male rats (Sprague-Dawly, 4-5 weeks of age) was sectioned under pentobarbital-Na anesthesia. The extensor digitorum longus muscle and the soleus mus cle were isolated 10-21 days after the operation and were suspended in a 10 ml bath of mammalian Ringer's solution according to Isaacson et al (3), which was aerated with a mixture of 95 %02 and 5 % CO2 and kept at 37-1-1°C (pH 7.5). In order to examine twitch response, field stimulation was applied to the muscle by platinum electrodes with the rec tangular pulses of suprarnaximal voltage and 5 cosec duration at 0.1 Hz frequency. The initial tension of the muscles was maintained within the range of 0. 1-0.3 g. Semi-isometric responses of twitch and resting tension were recorded on an ink-writing oscillograph through a strain gauge. Drugs were added to the bath in 0.1 ml volume to make the final required concentrations. EPI, norepinephrine (NOR) and trimetoquinol (TMQ) were used as sympathomimetics. TMQ is known to possess more potent (l-stimulating activity than isoproterenol in the tracheal muscle of guinea-pigs (4, 5), and the drugs also showed only a fl-action in the skeletal muscles of rats and rabbits (2, 6), while isoproterenol had a weak a-action besides (l-action (6).EPI in a concentration of 10-' g/ml caused an increase in the resting tension (con tracture) of the chronically dencrvated extensor digitorum longus muscle, as shown in Figs. I and 2. Higher concentration (10-6 10-5 g/ml) of the drug produced more remark able contracture with a steeper slope of rising phase and a larger peak of the tension, while lower concentration (10-e g/ml) sometimes failed to elicit the effect. This contractural response was differentiated into two components, a fast phase and a slow phase, which were sometimes observed very clearly as shown in Fig. 1-C. NOR produced similar con tracture to that by EPI_ TMQ caused only a slow phase of the contracture, the onset of