Binding of cardiac steroids to the Na,K-pump may occur at both high- and low-affinity sites. Binding at low-affinity sites causes pump inhibition. However, the functional significance of binding at high-affinity sites is controversial. The effects of cardiac steroids presumed bound at high-affinity sites on electrogenic Na-pumping was therefore examined in human atrial tissue. Specimens were preloaded with Na+ by cooling to 2 degrees C for 60 min and then rewarmed to 30 degrees C in 20 mM K+. In nine control experiments, the resting membrane potential hyperpolarized to -69.9 +/- 0.7 mV (mean +/- S.E.) due to enhanced Na,K-pump activity upon rewarming. Maximal hyperpolarization was less negative during exposure to 10(-9) to 10(-6) M acetylstrophanthidin in a dose-dependent manner. The differences between maximal hyperpolarization during control experiments and during exposure to acetylstrophanthidin were attributed to pump inhibition. Inhibition occurred at all acetylstrophanthidin concentrations including the low range in which binding at high-affinity sites should predominate. The relationship between acetylstrophanthidin concentration and pump inhibition was bimodal with a "shoulder" around 10(-8) M. The data were well described by a model which assumes that acetylstrophanthidin binds at both low capacity-high-affinity (half-saturation, 3.9 X 10(-9)M) and high capacity-low-affinity sites (half-saturation, 4.8-6.1 X 10(-7) M). Nonlinearity in the experimental assay of pump function did not appear to explain the bimodal concentration-inhibition relationship. Thus, cardiac steroids may cause Na,K-pump inhibition in humans by binding to high-affinity pump sites.