The soleus muscle of most animals exhibits a slow contraction and contains principally fibres which are low in myofibrillar ATPase but rich in oxidative enzymes (type I). The motor unit compositions of cat (McPhedran, Wuerker & Henneman, 1965) and rat soleus (Close, 1967) reflect these properties. In contrast the mouse soleus contains a high proportion (about 60 %) of type II fibres and its contraction speed is intermediate between that of typically fast muscles, such as rat EDL, and pure slow muscles, such as cat or guinea-pig soleus. We have, therefore, examined the motor unit profile of soleus muscles from normal mice of the C57 B1/6J strain. Mice were anaesthetized with sodium pentobarbitone, 60 mg/kg i.r. A fine silicone rubber tube was sealed into the peritoneal cavity with cyanoacrylate adhesive, and used to maintain anaesthesia. Units were isolated by splitting ventral root filaments in the conventional manner. Immediately after the laminectomy, mice weighing about 25 g were given an infusion of 0 5-1 0 ml. of whole citrated blood. This was found to prolong survival time and to improve the condition of the preparation. In forty-seven motor units tetanic tensions, expressed as a percentage of whole muscle tension, showed little variation, ranging from 1-20 to 9-6%. The mean value was 4-60 % suggesting that soleus contains about twenty-two motor units. Time to peak tension showed a much greater variation, ranging from 7 to 37 msec in ten solei with whole muscle times to peak between 13 1 and 21-6 msec. No clear division into two groups of fast and slow units, as observed in rat soleus by Close (1967), was apparent in mouse soleus. As also observed in the cat (Bagust, 1974) in individual muscles a linear relationship was found between axonal conduction velocity and time to peak tension. Mouse soleus contains some motor units with times to peak as short as those in rat fast muscle and others almost as slow as those in rat soleus (Close, 1967).