Posttraumatic stress disorder (PTSD) is highly prevalent, difficult to treat, and has very serious consequences for veterans and their families. Exposure therapy is commonly used and has positive results in a large percentage of cases, but improvements in therapeutic effects are sorely needed. We propose a novel strategy that uses a combination of state-of-the-art technologies to develop an improved exposure therapy approach. We plan to use a virtual reality experience (Virtual Iraq, in which participants will be virtually driving through Iraqi landscapes where different stressors will be present, such as explosions, enemy combatants, etc.). Virtual reality has already been shown to improve on regular exposure therapy. The main innovation is the use of real time functional MRI (fMRI) during the virtual reality exposure. We will first collect brain images under two clearly defined states: CALM (driving though Iraq with no stressors or war-like scenes) and STRESS (war-like Iraq). We will then use a machine learning algorithm to define those two brain states in terms of neural activity. Next, we will expose participants to a virtual Iraq environment in which they will be able to modify the scenes from stressful to calm, by bringing their brain state to CALM. A real time signal in the screen will let them know where in the continuum CALM-STRESS their brain is at each moment. We hypothesize that this feedback signal will allow for the participant to learn which mental strategies (think pleasant thoughts, concentrate on breathing, etc.) work best in terms of calming the brain. Once the brain moves to CALM, the Virtual Iraq scene will also move to less stressful, providing additional feedback. We propose that this training (two sessions to study individual CALM/STRESS states, 6 sessions of neurofeedback training) will allow patients to develop their own personalized mental strategies to be able to calm their brain in real life, when exposed to triggers. A second innovation is the use of a wearable device to continuously collect physical activity, heart rate, and sleep quality data throughout the 9 weeks of the study. We hypothesize that these measures may result in data more reliable than personal reports during treatment. In conclusion, we propose to use neurofeedback in a virtual reality environment to study the possibility that these real time signals will help veterans with PTSD develop individualized mental strategies to become calm when under stress. Since MRI is very expensive to use as a common intervention, we propose that once we demonstrate the feasibility of the approach on this pilot project, other less expensive and easier to use techniques (perhaps EEG) can be developed based on our data.