Specific [3H]ouabain binding to rat and guinea pig skeletal muscle (musculus soleus) was studied using a rapid centrifugation and a filtration method. Both assays gave identical results: the incubation of the cell membranes in 50 mM imidazole/HCl buffer pH 7.25 or 7.4 MgCl2, Pi caused a time dependent loss of (Na+ +K+)-ATPase activity indicating an alteration of the membrane preparation. Ouabain binding properties were changed concomitantly. If ouabain binding was allowed to proceed until equilibrium was reached (3 min in rat and 10 min in guinea pig) at 37 degrees C the data plotted according to Scatchard followed a straight line. The dissociation constants of the ouabain-receptor-complexes of the rat cell membrane preparation as calculated from the slope of the plot (KD = 134 nM) and from the ratio of the dissociation and association rate constants (KD = 175 nM) agreed within experimental error with that determined by Clausen and Hansen [(1974) Biochim. Biophys. Acta 345, 387-404] in intact soleus muscles (KD = 210 nM). If ouabain binding was allowed to proceed for a longer period, however, nonlinear Scatchard plots resulted with an identical maximal number of binding sites but inconstant and decreased affinity for the cardiac glycoside. Experimental evidence is presented that nonlinear Scatchard plots often obtained in hormone (drug)-receptor binding experiments may (among other things) be the result of damaged cell membrane particles in vitro.