This paper evaluates the feasibility and reliability of measuring the (emotional) state of the robot operators in urban search and rescue missions in real-time. An experiment has been conducted, in which a high-fidelity team task in a realistic urban search and rescue setting was executed by fire fighters in cooperation with robots. During the task, several emotion-eliciting events were triggered. In addition, the heart rate variability, skin conductance and facial expressions were monitored. After the scenario execution, the fire fighters were asked to describe their emotional state during task execution. We found that the facial expressions were not reliably recognized, but that heart rate variability and skin conductance measured a higher arousal during (some of) the emotion-eliciting events. However, the different measures still have shortcomings regarding use in complex and dynamic environments. © 2013 ACM.