Untreated septic shock results in depletion of extracellular fluid, cellular swelling, increased intracellular sodium, and decreased intracellular potassium concentrations in primate skeletal muscle. The Langer rabbit heart interventricular septal preparation was used to determine whether similar changes occur in cardiac muscle during sepsis. Rabbit septa (n = 17) were perfused with control and septic rabbit plasma plus red blood cells. Tissue contractility (developed tension [DT] and rate of tension change [dP/dt]) was followed, plasma cations were measured (Na+, K+, Ca2+, H+), perfusion pressure (PP) was monitored, and 42K efflux was determined. The effect on 42K efflux caused by the addition of potassium chloride to control plasma was determined. During perfusion with septic plasma there was significant decline of septal function (P less than 0.001). In 12/17 experiments DT fell 77.8 +/- 21.4% and dP/dt fell 75.8 +/- 24.8% from control values (means +/- 1 SD). All septa recovered when perfusion with control plasma was resumed. If [K+] was increased in control plasma during 42K washout, the percentage increase of effluent counts per minute per minute correlated with the percentage rise of control plasma [K+] (r = 0.95, P less than 0.001). During perfusion with septic plasma there was no similar correlation (r = 0.277). 42K efflux increased during septic plasma perfusion independent of the differences between control and septic plasma [K+], demonstrating abnormal myocardial K+ efflux. An abnormal efflux of K+ is seen during septic plasma perfusion similar to that described in primate skeletal muscle. It is associated with and may be a mechanism of action for the observed fall of contractility.