Preventive measures to reduce the risk of decompression sickness can involve several procedures such as oxygen breathing during in-water decompression. Theoretical predictions also suggest that brief periods of recompression during the course of decompression could be a method for controlling bubble formation. The aim of this study was to get clearer information about the effects of different experimental ascent profiles (EAPs) on bubble reduction, using pure oxygen or recompression during decompression for nitrox diving. Four EAPs were evaluated using bubble monitoring in a group of six military divers using Nitrox 40% O(2) breathing with a rebreather. For EAP 1 and 2, 100% O(2) was used for the end stage of decompression, with a 30% reduction of decompression time in EAP 1 and 50% in EAP 2, compared to the French navy standard schedule. For EAP 3 and 4, nitrox 40% O(2) was maintained throughout the decompression stage. EAP 3 is based on an air standard decompression schedule, whereas EAP 4 involved a brief period of recompression at the end of the stop. We found that EAP 1 significantly reduced bubble formation, whereas high bubble grades occurred with other EAPs. No statistical differences were observed in bubbles scores between EAP 3 and 4. One diver developed mild neurological symptoms after EAP 3. These results tend to demonstrate that the "oxygen window" plays a key role in the reduction of bubble production and that breathing pure oxygen during decompression stops is an optimal strategy to prevent decompression sickness for nitrox diving.