A dominant conceptualization of empathy in social work practice and education, provided by Karen Gerdes and Elizabeth Segal, relies heavily on the simulation theory adopted directly from the cognitive neurosciences. The aim was to critically challenge such a view by reporting on some recent empirical findings from the field in which professional social workers were interviewed about their experience of empathy in working with forced migrants. The findings support a phenomenological, critical account of simulation theory and provide evidence that empathy is constituted as a direct social perception of the other's experience. It was concluded that simulation theory is insufficient in providing an exhaustive approach for the professional use of empathy in social work practice and education.