Rationale: Andersen-Tawil syndrome (ATS) represents ventricular arrhythmia with abnormal U wave, periodic paralysis, and dysmorphysm, resulting from reduction of I K1 current due to KCNJ2 gene mutation. There are few reports about prognosis of ATS. In 2013, a French group reported that the prognosis of ATS was relatively benign under treatment. However, some our KCNJ2 -positive patients show aborted cardiac arrest in their adult phase. Here, we examined long-term outcome in Japanese KCNJ2 -positive ATS cohort. Methods and Results: In 32 probands with at least one of ATS features, we screened KCNJ2, KCNQ1, KCNH2 , and SCN5A genes. Excluding 2 compound mutations and 1 double mutation cases, we found 14 KCNJ2 mutations in 23 probands (71%). We sent the inquiry sheet to all the attending physicians of probands and collected probands’ symptoms, ECGs, and history of cardiac events. For the average follow-up period of 79 months, four of 23 probands (17%) experienced syncope. Twenty-two of 23 (96%) patients showed ventricular arrhythmias, 10 patients (43%) showed periodic pararysis. We compared 23 probands dividing into two groups; one with syncope (n=4), another without syncope (n=19). There are no significant difference in gender, patient number of ventricular arrhythmia, periodic pararysis, and position of KCNJ2 mutations. ECG findings such as HR (61bpm vs. 80 bpm), QTc (439 msec vs. 404 msec), QUc (668 msec vs. 650 msec), and Tpeak-Upeak interval (233 msec vs. 198 msec) showed no significant difference. The age of diagnosis was significantly higher in syncope group (32 ± 9 year old vs. 16 ± 10 year old, P=0.014). Aborted cardiac arrest (n=2, 50% vs. n=0, P=0.0046), beta blocker use (n=4, 100% vs. n=7, 37%, P=0.011) and ICD implantation (n=2, 50% vs. n=0, P=0.0046) were significantly higher in syncope group. Conclusion: We need to watch attentively KCNJ2 positive probands even after childhood.