In general, the building environment is a system of unsteady variables that are constantly changing (or fluctuating), even within a single space. Current monitoring capabilities such as thermostats do not fully capture the detailed dynamically changing processes for major environmental parameters in a building/space with regard to both spatial and temporal variations. Detailed indoor environmental information can be provided for potentially better building environmental control and building energy performance improvement using fast calculations from a proper orthogonal decomposition-linear stochastic estimation (POD-LSE) methodology with limited measurements. In this work, the POD-LSE method has been used to reconstruct the entire fluctuating thermal field in an experimental room with a high level of accuracy (error less than 0.56 degrees C for cooling) based on a few real-time measurements. Evaluations to the mechanically conditioned thermal environment through a single temperature from the thermostat and the comprehensive mean temperature of the occupied zone from the proposed POD-LSE method have been compared. The dynamic PMV values from reconstructed temperatures and reconstructed air movement velocities based on measurements only at the supply diffuser and the thermostat location are also presented, which show a good agreement with the PMV values calculated directly from measurements. Published by Elsevier B.V.