SUMMARY1. Rats were given drinking water containing either 20 mm-CsCl or 20 mM-RbCl for a period of 2 weeks. Samples of blood were then taken from the rats under anaesthetic. They were immediately centrifuged and the plasma taken for analysis. Soleus muscles, diaphragm, extensor digitorum longus, white gastrocnemius and vastus lateralis muscles were then taken from the dead animals and these and the plasma were analysed for potassium, and for caesium or rubidium by means of the flame photometer.2. The concentrations of potassium and rubidium or caesium in the fibre water of these various muscles and in the samples of plasma water were then calculated.3. It was found that the red muscles including soleus and diaphragm generally tended to accumulate caesium and rubidium to a greater extent than did the white muscles such as the white gastrocnemius and vastus lateralis. 6. When these values in the caesium-treated animals were plotted against the percentage of red fibres in the five muscle types (as obtained from the data of Sreter & Woo, 1963) the graph indicated that the white fibres had similar ionic gradients for Cs+ and K+ and that affinity for Cs+ was confined to the red fibres.7. The membrane potential measured in soleus and extensor muscles immersed in plasma from the same animal was not significantly different from EK but was much less than Ecu.7-2 8. These results are interpreted in terms of permeability differences between the slow red fibres and white twitch fibres.