BACKGROUND The effectiveness of virtual reality exposure therapy (VRET) in the treatment of anxiety disorders is well established. Several working mechanisms of VRET have been identified, whereby both emotional processing and the sense of presence play a key role. However, there are only few studies that contribute to our knowledge on the working mechanisms in the case of VRET for claustrophobia. OBJECTIVE The current study asks for the experiences and perspectives of patients, healthy subjects and therapeutic experts testing a VRET application with 5 intensity levels for symptoms of claustrophobia in order to detect working mechanisms and find suggestions for improvement. METHODS A mixed-methods design was applied in which patients (n=15), healthy subjects (n=15) and therapeutic experts (n=15) tested a VRET intervention of an elevator ride in 5 intensity levels. Intensity was varied by elevator-size, duration of elevator-ride, and presence of avatars. Quantitative measures examined pre-treatment state-anxiety (STAI-S), the technology commitment scale (TCS) and post-treatment self-reported presence with the igroup presence questionnaire (IPQ). Finally, a quality score was calculated on the basis of 15 Likert-scaled evaluation items that had been developed for the purpose of this study ranging from 0 to 5. Think-aloud protocols of the patients and semi-structured interviews post-treatment of all participants were conducted to gain in-depth perspectives on emotional processes. RESULTS Pre-treatment anxiety was highest in the patient group with a STAI-S score of M=45.79 (healthy subjects: M=35.47; experts: M=36.80). Technology commitment was M=4.02 among all participants. Post-treatment scores showed a high IPQ presence score (M=3.88, SD=0.90) with differences between the three groups (P=.03*), especially lower values for patients (M=3.77, SD=0.91). Overall evaluation was high with M=4.24 (SD=0.35). The think-aloud protocols of the patients revealed that anxiety and presence both were achieved. Qualitative interviews uncovered 8 topics including feelings & emotions, personal story, telepresence, therapeutic effects, barriers, conditions and requirements, future prospects and realization. CONCLUSIONS The VRET intervention for claustrophobia evoked anxiety in patients and tension in the controls. A feeling of presence was reported by all participants. Patients expressed to prefer the presence of a therapist while doing the intervention. Experts rated the intervention as clinically valuable. Due to the implementation of avatars some of the fears might be related to symptoms of social phobia or agoraphobia. A higher degree of sophistication in the intensity levels is needed to deliver targeted help for specific symptoms of anxiety.