The short- and long-term outcomes of patients treated with class I antiarrhythmic drugs were studied. A total of 968 patients who began class I antiarrhythmic therapy in our hospital were followed up for a mean of 13 +/- 8 months. Discontinuation rates were similar for the entire group of patients during short-term therapy (52%) and long-term therapy (50%). During short-term therapy, 28% of patients stopped treatment because of side effects, and 21% of patients discontinued treatment because of ineffectiveness. Discontinuation rates during short-term therapy were significantly lower with encainide, flecainide, and procainamide than with the other agents (disopyramide, mexiletine, quinidine gluconate, quinidine sulfate, and tocainide) (P < 0.05). However, encainide, flecainide, and procainamide were associated with a higher incidence of side effects necessitating discontinuation in the long-term study. Side effects during long-term therapy accounted for 27% of discontinuations, while loss of efficacy accounted for only 6%. Discontinuation rates were significantly lower with mexiletine, quinidine gluconate, quinidine sulfate, and tocainide than with the other agents during long-term therapy (P < 0.05). The discontinuation rate with quinidine gluconate was significantly less than that with quinidine sulfate. All antiarrhythmic agents possess a substantial potential for toxicity. Short-term response and tolerance to class I antiarrhythmics do not predict the ultimate response to therapy. Thus patients receiving class I antiarrhythmic therapy must be monitored for the duration of treatment.