We have previously shown reduced selectivity for depolarized and acidotic myocardium for encainide and flecainide compared to lidocaine and amiodarone. The present study aims to compare propafenone and two of its metabolites (5-OH-propafenone and N-despropylpropafenone) to lidocaine in the same model. Standard microelectrode methods were used to record intracellular action potentials from strips of guinea pig right ventricular myocardium superfused with either standard physiological saline (pH 7.3; pO2 > 600 mm Hg; [K+] = 5.6 mM), or the same solution modified to produce either hyperkalemia (K+ = 11.2 mM), acidosis (pH = 6.3), or hypoxia (pO2 = 60 mm Hg). The effects on action potential parameters of three ''therapeutic'' concentrations of lidocaine, propafenone, and two of its metabolites were studied under all four conditions at four different drive rates from 200 to 25 beats/min. Hyperkalemia, in the absence of drugs, produced reductions in the resting potential (86. 7 +/- 2.5 to - 71.8 +/- 3.7 mV) and the maximum rate of depolarization (V(max), 300.0 +/- 46.5 to 205.6 +/- 37.6 V/s). All four drugs produced increased depression of V(max) in hyperkalemia and acidosis compared to control conditions, but it was a consistent finding that at concentrations that were approximately equipotent in control conditions, lidocaine produced greater increments in depression of V(max) in hyperkalemic and acidotic superfusate than did propafenone or either of its metabolites. Qualitatively similar results were obtained for both metabolites compared to lidocaine. Hypoxia produced no significant modulation of drug effects. We conclude that propafenone and its metabolites behave like the other class Ic drugs, encainide and flecainide, in the presence of some of the major ''components'' of ischemia. The fact that lidocaine, but not propafenone, can be given in concentrations that suppress ''ischemic'' myocardium but have little effect on normal tissue may be of relevance to the proarrhythmic potential of propafenone.