For the first time, response of personal exposimeters (PEMs) is studied under diffuse field exposure in indoor environments. To this aim, both numerical simulations, using finite-difference timedomain method, and calibration measurements were performed in the range of 880-5875 MHz covering 10 frequency bands in Belgium. Two PEMs were mounted on the body of a human male subject and calibrated on-body in an anechoic chamber (non-diffuse) and a reverberation chamber (RC) (diffuse fields). This was motivated by the fact that electromagnetic waves in indoor environments have both specular and diffuse components. Both calibrations show that PEMs underestimate actual incident electromagnetic fields. This can be compensated by using an on-body response. Moreover, it is shown that these responses are different in anechoic chamber and RC. Therefore, it is advised to use an on-body calibration in an RC in future indoor PEM measurements where diffuse fields are present. Using the response averaged over two PEMs reduced measurement uncertainty compared to single PEMs. Following the calibration, measurements in a realistic indoor environment were done for wireless fidelity (WiFi-5G) band. Measured power density values are maximally 8.9 mW/m 2 and 165.8 mW/m 2 on average. These satisfy reference levels issued by the International Commission on Non-Ionizing Radiation Protection in 1998. Power density values obtained by applying on-body calibration in RC are higher than values obtained from no body calibration (only PEMs) and on-body calibration in anechoic room, by factors of 7.55 and 2.21, respectively. Bioelectromagnetics. 2016;9999:XX-XX.