Studies have shown that continuous rhythm monitoring enables the detection of significantly more atrial fibrillation (AF) episodes than routine follow-up of patients, i.e. based on perception of symptoms or on 24-48 h Holter monitoring. The positive outcome of radiofrequency ablation (RFA) may be easily overestimated, especially in patients with paroxysmal AF. Thirty-three consecutive patients, aged 59.4+/-8.9 years (range 38-75 years) participated in this study. All patients had documented AF episodes with an AF duration of 9.4+/-7.1 years (range 1.5-25 years). A new monitoring device, the AF-Alarm was used to more accurately assess the outcome after surgical isolation of pulmonary veins. The AF-Alarm was applied for a duration of 128+/-42.5 h (range 49-191 h) during a period of 8-15 days. The success rate was 87% based on serial electrocardiograms (ECGs) and 24-48 h Holter monitoring during regular outpatient visits. Combination of ECG, Holter and AF-Alarm data yielded a significantly lower success rate, i.e. at the latest follow-up 69% of the patients were free from AF after surgical ablation (P<0.05). Furthermore, the AF-Alarm device demonstrated a dissociation between symptoms and atrial arrhythmic events and confirmed the occurrence of asymptomatic AF episodes. The most important limitation of the AF-Alarm device was noise detection with oversensing and inappropriate detection of non-existing AF episodes in 9% of patients. Long-term follow-up of the patients seems to be essential as success rates of the initial ablation procedure might vary over time. External recorders like the AF-Alarm may be used as an additional tool to document symptomatic and asymptomatic episodes of atrial arrhythmias in the outpatient setting.