PURPOSE: The dynamics of the postulated phenomenon of exercise baroreflex resetting is poorly understood, but can be investigated using closed-loop procedures. To shed light on some mechanisms and temporal relationships participating in the resetting process, we studied the time course of the relationship between the R-R interval (RRi) and arterial pressure with a closed-loop approach. METHODS: On ten young volunteers at rest and during light exercise in supine and upright position, we continuously determined, on single-beat basis, RRi (electrocardiography), and arterial pressure (non-invasive finger pressure cuff). From pulse pressure profiles, we determined cardiac output (CO) by Modelflow, computed mean arterial pressure (MAP), and calculated total peripheral resistance (TPR). RESULTS: At exercise start, RRi was lower than in quiet rest. As exercise started, MAP fell to a minimum (MAPm) of 72.8 ± 9.6 mmHg upright and 73.9 ± 6.2 supine, while RRi dropped. The initial RRi versus MAP relationship was linear, with flatter slope than resting baroreflex sensitivity, in both postures. TPR fell and CO increased. After MAPm, RRi and MAP varied in opposite direction toward exercise steady state, with further CO increase. CONCLUSION: These results suggest that, initially, the MAP fall was corrected by a RRi reduction along a baroreflex curve, with lower sensitivity than at rest, but eventually in the same pressure range as at rest. After attainment of MAPm, a second phase started, where the postulated baroreflex resetting might have occurred. In conclusion, the change in baroreflex sensitivity and the resetting process are distinct phenomena, under different control systems.