SUMMARY Limited information is available on thermal comfort performance of the indoor environment in health care facilities both for staff and patients. Thermal comfort models such as Predicted Mean Vote (PMV) and Adaptive Thermal Comfort (ATC), have not been applied extensively for this setting. In this research these models have been tested. Thermal comfort perception and measurements have been performed in three hospital wards and over different seasons. Staff activity level was monitored in more detail to support the analysis. Results show large differences between PMV and AMV for staff members. ATC does not seem able to capture staff thermal comfort in detail. Better agreement is found for patients. To better align with the actual conditions and individualized activities in hospitals, reference to new biophysical thermal comfort models might be an interesting alternative.