Although immersive virtual environments (IVEs) have been increasingly applied to building design, they mostly provide audio and visual perception to users, but not thermoception. This is a significant limitation, because thermal state (i.e., thermal sensation, thermal comfort, and thermal acceptability) plays an important role in understanding human-building interactions during design. Mixed immersive virtual environments (MIVEs), i.e., combining IVEs with a climate chamber, are a potential solution to thermoception analysis during building design. Experiments were conducted to test the hypothesis that participants’ virtual experience did not significantly alter their thermal state compared to their in-situ experience. Response variables were the control temperature distribution over thermal state scales, the thermal state vote distribution over thermal state scales at a certain temperature step, and the physiological responses. The results show that the first two response variables were not significantly different between the in-situ and the MIVE experiments. Due to wearing the head mounted display in the MIVE experiments, the forehead skin temperatures were significantly higher than those in the in-situ experiments in most cases. However, such difference in skin temperatures did not affect the general thermal state. MIVEs are potentially a reliable apparatus to collect thermal state data for design decision-making.