BACKGROUND: A randomized double-blind monocentric trial was conducted at the spa of Royat, France, in patients with mild Raynaud's phenomenon (one or two attacks a day during the autumn and winter months). Hemodynamic changes were quantified by laser Doppler flow (LDF) and computerized chronothermometric test (CT) during local application of CO2 and compared with the usual clinical parameters. PATIENTS AND METHODS: Patients were randomized to 18 days of CO2 (group 1) or 9 days of air + 9 days of CO2 (group 2). The gas was applied at a constant temperature to previously humidified skin on the forearm. Effects of the active treatment were measured by LDF and continuing benefits by computerized chronothermometric test (CT test). Clinical assessment was made on four occasions (before, in the middle and at the end of treatment, and three months afterwards) on the basis of the number of daily attacks and the duration and severity of the attacks. RESULTS: LDF showed a significant increase in digital blood flow (+ 41%) and in vasomotion (+ 42%) during CO2 treatment but no change on placebo treatment. CT test showed a comparable response to cold in the two groups during treatment. Three months later, however, warming time was significantly decreased (by 7 minutes: 25%) and warming rate increased (+ 22.4%) in group 1 (intergroup difference: p = 0.02). There was no difference between groups in the weekly number of attacks during treatment periods. In contrast there was a moderate increase in both groups in the winter months. There was no increase in the severity of the attacks nor in their duration. CONCLUSION: In this clinical trial, LDF evidenced positive effects of percutaneous application of natural CO2 gas on vasomotricity and vasomotion while CT test showed that patients who had received 18 days of CO2 adapted more easily to exposure to cold. Clinical improvement seems to be masked by winter weather conditions.