BACKGROUND: Mean heart rate and irregularity of the rate, i.e., heart rate variability (HRV), are two aspects of heart rate during atrial fibrillation (AF). An important goal of AF therapy is to control mean heart rate during exercise; the determinants of HRV during AF remain poorly known although its prognostic value has been established. OBJECTIVES: To investigate the effects of a stable, long-acting, selective A1-adenosine receptor agonist, SDZ WAG994, on heart rate during exercise and on HRV. METHODS: In a multicenter, double-blind, randomized, placebo-controlled, parallel group study, patients with permanent AF performed a symptom-limited exercise test and underwent 24-hour ECG monitoring on day 1 during treatment with placebo, and on day 2 during treatment with either placebo or 2 mg SDZWAG994 orally. Changes in mean heart rate during exercise and changes in HRV indices between day 1 and day 2 were compared between the two groups. RESULTS: Thirty-two patients (64 +/- 8 years; 81% male; 25% in NYHA Class II; 38% with no structural heart disease) were included in the study. During active treatments, heart rate remained unchanged at rest and increased significantly during exercise. A significant daytime increase in short-term HRV indices (DpNN50 = 4.5% P = 0.01; DrMSSD = 6% P = 0.03; DSDNN Index = 6% P = 0.02) occurred during active treatment. CONCLUSIONS: Selective A1-adenosine receptor agonism with SDZ/WAG994 limits the increase in mean heart rate during exercise in patients with AF. In addition, this agonist selectively increases short-term HRV indices, suggesting that pNN50, rMSSD, and SDNN reflect vagal influences during AF.