Background: Empathy is an individual's capacity to understand the behavior of others, to experience their feelings, and to express that understanding to them. Empathic ability is an asset professionally for individuals, such as teachers, physicians and social workers, who work with people. Being empathetic is also critical to our being able to live with others in general, and ultimately to leading happier lives. Subsequently it seems imperative to examine if and how it is possible to enhance people's empathic ability. Purpose: The purpose of this article is to use narrative review method to analyze studies of empathy training in human service and social science disciplines over the past thirty years to address the questions: "How have people been trained in empathy and what are the findings?" and "How was empathy training evaluated and how valid are these evaluation findings?" Setting: Not applicable. Intervention: Not applicable. Research Design: Not applicable. Data Collection and Analysis: Narrative review. Findings: Twenty-nine articles pertaining to empathy training evaluation research were identified based on an advanced computer search on the following databases: "Education Full Text," "ProQuest Education Journals," "Web of Knowledge" and "Educational Resources Information Center" (ERIC). Seven types of training methods were noted in these 29 evaluations with the most popular being didactic related (42%). All but two studies (93%) reported positive findings, mainly in regard to learning (86%), or the cognitive component of empathy. These findings suggest that regardless of the training method, individuals can learn about the concept of empathy. Unfortunately, information pertaining to the effects of training on individuals' feeling for others, and their ability and propensity to take the perspective of others and to demonstrate it in the natural environments is lacking. Consequently, very little is known about the trainability of the affective and behavioral components of empathy. Also, some of the findings were moderated by gender, age, education level, and time of measurement. Regarding evaluation research designs, most of the studies used self-reporting to collect trainees' knowledge about empathy and most of the quantitative studies used a control group and pretesting to examine training impact. Construct validity of both empathy measurement and training is very problematic. A majority of the studies did not clearly define empathy, provide training as defined, and/or measure what is being trained; conceptualization of empathy across studies was not consistent either. In sum, data from the studies reviewed were neither complete nor valid enough to provide a clear and full understanding of the trainability of empathy. More research is apparently needed and hopefully lessons learned from our review will be considered in designing future studies. (Contains 9 figures and 1 table.)