Nitrous oxide scavenging devices have not been tested in the prehospital setting. We studied two nitrous oxide scavenging devices and their effect on ambient levels of nitrous oxide in the patient compartment of a stationary ambulance. We compared the ambient levels of gas when nitrous oxide scavenging was in use versus when it was not. In addition, we evaluated the effect of two different mask designs on ambient levels of nitrous oxide. With eight healthy male volunteers acting as their own controls using a series of mask/scavenger combinations, we found that there was a significant difference in the ambient levels of nitrous oxide when a scavenger was in use. The mean time-weighted average was 776 parts per million/min when no scavenger was used. When gas scavenging was in use, the mean time-weighted average never exceeded 150 parts per million/min. (P less than .0001). There was no significant difference between nitrous oxide scavengers or between the masks tested in their effect on nitrous oxide ambient levels in the atmosphere. We conclude that nitrous oxide scavengers are effective in lowering nitrous oxide ambient levels in the prehospital setting and merit consideration as adjuncts to nitrous oxide delivery systems in this setting.