Virtual reality (VR) is characterized by a high sense of immersion. Objective and quantitative evaluation methods are essential for enhancing the immersiveness of VR technology. The sense of immersion can be represented by attention allocation from the real world to the virtual space. Our previous study (Ogawa et al.: Adv Biomed Eng. 11, 1-9, 2022) proposed a unique approach that utilized auditory P300 responses to measure attention allocation during 2D or 3D video watching. Using sound stimuli, attention allocation from the real world to the virtual space was quantitatively evaluated using P300 amplitudes. However, this method was limited to assessing immersion in VR contents without audio because sound was used as a probe stimulus. In the present study, we adopted the somatic probe stimulus method to quantitatively and objectively evaluate attention allocation while watching audio‒VR videos. We developed a novel somatic P300 system that involves the application of vibration stimuli to the user’s fingers during VR video viewing. Ten young adult participants performed a somatic oddball task while experiencing the VR content. The small amplitude of the event-related P300 wave during the oddball task quantified the attention directed towards the VR content. Vibrations of the piezoelectric device, which were delivered randomly as probe stimuli, were applied to the thumb (standard stimulus, 70%), middle finger (target stimulus, 15%), and little finger (deviant stimulus, 15%). EEG signals were recorded at Cz and Pz, filtered from 0.5 to 30 Hz, with a sampling frequency of 1000 Hz. To minimize electromyogram signal contamination, mental counting replaced the switch-pressing method used in the previous study. Event-related potential waveforms obtained by averaging signals from ten subjects revealed distinctive P300 responses for target stimulus and deviant stimulus. The peak amplitudes of the P300 wave for target and deviant stimuli during VR video viewing were significantly smaller than those during no VR video viewing. These results suggest that the proposed method can be used for objective and quantitative measurements of immersion intensity, even in VR environments with audio.