The flecainide test is widely used in Brugada syndrome. However, its reproducibility and safety remain ill-defined. This study included 22 patients (18 men, mean age 34 years). Mutations in the SCN5A gene were found in eight patients. Two patients had aborted sudden cardiac death, 8 had syncope/presyncope, and 12 were asymptomatic. The ECG was diagnostic in 19 patients and suggestive in 3. At baseline, 21 of 22 patients underwent a flecainide test (2 mg/kg IV bolus over 10 minutes). In 21 of 21 patients the test was diagnostic or amplified the typical ECG pattern. At the end of drug infusion, sustained VT lasting 7-10 minutes developed in two patients. A second flecainide test was performed within 2 months in 20 patients. The test was not repeated in the two patients with prior development of VT. The flecainide test was diagnostic in 20 of 20 patients. Sustained VT occurred in one patient and recurrent VF in another. The reproducibility of the flecainide test was 100%. In 4 (18%) of 22 patients major VAs were documented after the end of flecainide infusion. VA occurred in 3 (43%) of 7 patients with, versus 1 (7%) 15 without SCN5A gene mutation (P < 0.05). No diagnostic ECG changes or arrhythmias developed in 25 control patients without structural heart disease who underwent the same study protocol. This study shows a high flecainide reproducibility, supporting its diagnostic value in Brugada syndrome. However, the occurrence of major VA, significantly higher in patients with documented SCN5A gene mutation, including in asymptomatic patients, mandates the performance under appropriate medical supervision. Whether a slower rate of drug infusion can lower the risk of VA induction, while maintaining the sensitivity of the test should be explored.