1. Ring-demethylation of the pure antagonist bupranolol results in a ligand (K 105) which induces conformational beta-adrenoceptor changes leading to partial agonistic effects in heart and trachea. However, these conformational receptor changes are not accompanied by changes in receptor affinity, because the affinity estimates for K 105 and bupranolol did not differ for a variety of myocardial tissues (including ventricular beta-adrenoceptors labelled with 3H-(-)-propranolol] and trachea, not even for tracheal receptor subtypes. 2. For the analysis of the concentration-dependence of the blocking actions of a partial agonist a double log-plot was derived, which includes the classical Schild-plot as a special case. The plot is based on the statistical analysis of the action of partial agonists by Marano and Kaumann (1976). They defined a slope m for the weighted regression of equieffective concentrations of agonist in the absence and presence of a concentration [P] of partial agonist P. We derived the dependence of m on [P], which is suitably expressed as: log (1/m-1) = log [P]-log Kp. For the case of a single class of non-interacting receptors the slope of the double log-regression should be unity. Our plot has incorporated information from complete concentration-effect curves, instead of a single concentration-ratio as in the Schild-plot. Analysis of data of K 105 with the new plot (intrinsic activity greater than 0) and the Schild-plot (intrinsic activity = 0) yielded slopes near unity, consistent with simple competition.