For technical reasons, Rb-86 is frequently preferred to K-42 as a tracer for K+. Systematic comparisons of the two isotopes, however, are rarely done. In this paper we compare the transport of K-42 and Rb-86 in rat and mouse soleus muscle and in rat erythrocytes. Ouabain-suppressible KC uptake in rat soleus was the same whether measured with K-42 or Rb-86, both when stimulated by insulin, salbutamol and calcitonin-gene-related peptide (CGRP), and when inhibited by graded concentrations of ouabain. Control experiments with rat erythrocytes, where Na+-K+-Cl- co-transport has earlier been demonstrated, showed closely similar inhibitory effects of bumetanide on K-42 and Rb-86 uptake. In contrast, bumetanide produced dd significant change in K-42 uptake of rat and mouse soleus muscle, but clearly inhibited Rb-86 uptake at concentrations down to 10(-7) M (P < 0.001). Whereas the addition of 150 mM NaCl had no effect on K-42 uptake in rat soleus, Rb-86 uptake, and in particular the bumetanide-suppressible component, was markedly increased by this addition. The inhibitory effect of bumetanide on Rb-86 uptake gives rise to the false impression that skeletal muscle contains a NaKCl2 co-transport system. Efflux studies showed that the fractional loss of K-42 from rat soleus muscle is 2.3 times larger than that of Rb-86. Salbutamol and CGRP increased Rb-86 efflux, but inhibited K-42 efflux. This implies that for studies of K+ efflux and bumetanide-sensitive K+ transport, Rb-86 is not even an acceptable tracer for the detection of qualitative changes. Control experiments with K-42 are essential in any characterization of unknown K+ transport processes.