1 The positive inotropic responses of rat paced left atria to adrenaline and noradrenaline were recorded. Desmethylimipramine (DMI, 1 gM) and metanephrine (IO gM) were initially present throughout. 2 The positive chronotropic responses of spontaneously beating right atria to adrenaline were used as a reference. In these, pindolol, in increasing concentrations, caused progressive shift of the concentration-response curves to the right, which yielded a pA2 value (8.15) compatible with antagonism of /-adrenoceptors. 3 The left atrial tension responses to adrenaline showed an initial progressive displacement by pindolol (up to 3gM) which gave an unexpectedly low pA2 value (6.48). However, with further increases in pindolol concentration there was no additional shift of the curve. In the presence of pindolol (3 jM), prazosin (O.1 M) displaced the curve to the right but the pA2 value derived from this shift (7.75) was less than expected for ax-adrenoceptor antagonism. 4 When the experiments in the presence of pindolol (3 jM) were repeated in the absence of DMI, prazosin displaced the concentration-response curves for adrenaline-induced left atrial tension to a greater extent and the pA2 value (8.76) was now compatible with adrenaline stimulating typical a -adrenoceptors. 5 The concentration-response curves for noradrenaline-induced left atrial tension were also progressively displaced to the right by pindolol (0.1, 0.3 and 1.OpM). These concentrations yielded a Schild plot of unity slope and a pA2 value of 7.94 + 0.04. This was not significantly different from the pA2 value of 8.02 + 0.07 determined for pindolol against isoprenaline in the left atria, which indicates a normal interaction of noradrenaline with P-adrenoceptors in the absence and presence of low concentrations of pindolol.6 A further increase in the concentration of pindolol to 3 gM failed to induce an additional shift of the noradrenaline curves, whether a 'before and after' antagonist or a 'naive tissue' design was adopted. Similarly, the rightwards shift of the concentration-response curves by timolol reached a limit as the concentration was increased. In all cases the limit of shift occurred at a noradrenaline EC50 value of 5-1O pM. 7 At the limit of JJ-adrenoceptor antagonism, prazosin and dibenamine did not displace the noradrenaline curves further. The residual inotropic response to noradrenaline therefore appeared to be mediated via neither a-nor f-adrenoceptors. 8 DMI, in the absence of fl-blockade, produced the potentiation of adrenaline and noradrenaline expected of a neuronal uptake inhibitor. However, in the presence of pindolol, there was no potentiation of the right atrial rate response to adrenaline while its left atrial tension responses were antagonized. This suggested that DMI was acting as an a-adrenoceptor antagonist. It also explained the less-than-expected shift by prazosin of the adrenaline responses in the presence of both pindolol and DMI, the latter drug already exerting some a-blocking activity. In contrast, the left atrial...