Recent experiments on the pharmacological properties of the semisynthetic cardiotonic steroid strophanthidin-3-bromoacetate (SBA) have challenged the well-known potassium digitalis antagonism in isolated heart muscle preparations. In order to establish these results in vivo, the minimum lethal doses (LDmin) of ouabain (OUA), digoxin (DO), digotoxin (DT), k-strophanthidin (STR) and SBA were determined by the infusion toxicity method in guinea pigs at normokalemia and hypokalemia. The experimentally induced decrease of the serum potassium concentration (5.0 mmoles/l vs. 3.3 mmoles/l) significantly reduced the LDmin of DO (1.42 vs. 1.05 mumoles/kg), DT (1.78 vs. 1.24 mumoles/kg) and STR (20.16 vs. 15.98 mumoles/kg), whereas the LDmin of OUA (0.37 vs. 0.34 mumoles/kg) was not altered. Contrary, the LDmin of SBA was even slightly, but not significantly increased during hypokalemia (16.77 vs. 19.04 mumoles/kg). In addition, from the experimental data an optimum time of infusion (Topt), corresponding to the LDmin, can be derived, which is equivalent to the time for optimum "utilization" of the drug. The obtained sequence: STR less than OUA less than DO less than DT less than SBA represents the well-known differences in the onset of the pharmacological action in man resp. animal. Hypokalemia in general resulted in a shortening of the Topt, thus indicating a more rapid "utilization" of the drug tested. The above differences of the cardenolide action at reduced serum potassium concentration may be dependent on the recently reported divergent influence of potassium on the association- resp. dissociation rate constants for the interaction of these drugs with their binding site at the Na+-K+-ATPase.