We studied the effects of procainamide and N-acetylprocainamide (NAPA) in a conscious dog preparation of atrial flutter resulting from circus movement around the tricuspid orifice. We also recorded transmembrane potentials of atrial tissues from the circus path in vitro. In 12 instrumented dogs, average flutter cycle length was 157 msec, the duration of the excitable gap was 73 msec, and conduction velocity was 0.75 m/sec. At 4 and 8 mg/kg, procainamide moderately prolonged cycle length, but did not terminate the flutter. At a cycle length of 300 msec procainamide increased effective refractory period (ERP) December 1987 lamina.4 The reentrant excitation causing a flutter can be induced, reset, and terminated by premature stimulation. Furthermore, there is a partially excitable gap in the circus path and thus the rhythm can be entrained by programmed stimulation.3In this preparation of atrial flutter the path for circus movement is provided by electrophysiologically normal tissue4' 5; the reentrant impulse propagates in an anatomically discrete circuit and the arrhythmia can be studied over the long term in conscious animals. The preparation is thus suitable for study of the mechanism of action of antiarrhythmic agents. We do not argue here that flutter in our preparation results from the same mechanisms as flutter in the diseased human heart; indeed, in the latter different mechanisms probably operate in different cases. Nevertheless, our findings will be applicable to arrhythmias that result from the mechanism operative in our preparation. At