Aims: To study device performance, arrhythmia recurrence characteristics, and methods of outcome assessment using a novel implantable cardiac monitor (ICM) in patients undergoing ablation for atrial fibrillation (AF). Methods and results: In 419 consecutive patients undergoing first-time catheter ablation for symptomatic paroxysmal (n = 224) or persistent (n = 195) AF an ICM was injected at the end of the procedure. Telemedicine staff ensured full episode transmission coverage and manually evaluated all automatic arrhythmia episodes. Device detection metrics were calculated for ≥2, ≥6, and ≥10 min AF detection durations. Four methods of outcome assessment were studied: continuous recurrence analysis, discontinuous recurrence analysis, AF-burden analysis, and analysis of individual rhythm profiles. A total of 43 673 automatic AF episodes were transmitted over a follow-up of 15 ± 6 months. Episode-based positive predictive values changed significantly with longer AF detection durations (70.5% for ≥2 min, 81.8% for ≥6 min, and 85.9% for ≥10 min). Patients with exclusive short episode recurrences (≥2 to <6 min) were rare and their arrhythmia detection was clinically irrelevant. Different methods of outcome assessment showed a large variation (46-79%) in ablation success. Individual rhythm characteristics and subclinical AF added to this inconsistency. Analysis of AF-burden and individual rhythm profiles were least influenced and showed successful treatment in 60-70% of the patients. Conclusion: We suggest AF detection duration >6 min and AF burden >0.1% as a standardized outcome definition for AF studies to come in the future.