Emergency braking can generate forward displacement that may influence the effectiveness of protection in collisions, especially for passengers. The development of automated vehicles has enabled the diversification and rationalization of sitting positions, including reclined seating. However, the passenger response in pre-crash scenarios in reclined seating differs from that in standard seating, which poses different requirements for biofidelic human body models (HBMs) to evaluate passenger injuries in collisions. This study conducted emergency braking trials in vehicles at an initial velocity of 80 km/h. Five volunteers were exposed to approximately 1 g manual emergency braking (MEB), and the muscle responses at the front passenger seat with backrest angles of 25°, 45°, and 65° were recorded. The electromyography obtained from 14 muscles of the neck, torso, and lower extremity were normalized using maximum voluntary contractions (MVCs). In the quiet sitting phase, the activity levels were low (< 5% MVC) in all muscles for the three sitting positions. During emergency braking, the muscles are activated to restrict the body motion. There were differences in muscle amplitude and onset time in different backrest angles, with higher muscle activity levels in most muscles in a reclined sitting position. In particular, the sternocleidomastoid, rectus abdominis, and vastus medialis showed different patterns in the peak and steady-state phases. We found that the tibialis anterior was consistently activated at a lower level in all sitting postures (< 8% MVC), which indicates limited support of the shank for the body. The data provided in the paper are presented in corridors and intended for use in the development and validation of HBMs with active muscle models to simulate evasive maneuvers that potentially occur before a crash in the reclined sitting position.