This dissertation examines the question, how moral sensitivity can be promoted with digital games. Moral sensitivity alerts people to ethical problems, i.e., actions that could conflict with ethical norms, standards, values, and principles, or which have the potential to affect others’ wellbeing or the pursuit of their legitimate interests. The dissertation pursues two major research objectives. First, there is a lack of a unified theory on the nature of moral sensitivity, which integrates normative demands with empirical findings from various streams of research. A multitude of different ideas currently coexist on the nature of moral sensitivity. Prior efforts to integrate theories have been purely normative, ignoring contradictory empirical evidence. Empirically informed theories, on the other hand, tend to ignore findings that point to other explanations. The lack of an adequate theory of moral sensitivity stands in the way of systematic efforts to promote moral sensitivity, including through digital games. To fill this gap, this dissertation aims at proposing an empirically founded theory of moral sensitivity. Secondly, there is a lack of empirically informed recommendations on the use of digital games for the promotion of moral sensitivity. The major weaknesses of previous attempts to provide guidance on the promotion of moral sensitivity via digital games consist in (1) a lack of clarity about the goal of moral-sensitivity training, (2) a lack of knowledge on suitable strategies to promote moral sensitivity in general, and (3) a failure to take empirical findings on the effects of digital games on the development moral sensitivity and its components into consideration. This dissertation aims at overcoming these three limitations of previous research in proposing a first evidence-based framework for the promotion of moral sensitivity with digital games. The first part of this dissertation investigates the nature of moral sensitivity from a normative perspective. A critical analysis of the need for moral sensitivity suggests that its function is to alert people to ethical problems accurately, affectively, automatically, and autonomously. This position is elaborated in confrontation with alternative ideas, e.g., the belief that controlled deliberation should be involved in noticing ethical problems. The following examination of ethical demands on a good moral sensitivity from the perspectives of four paradigms of normative ethics suggests that three abilities are required: (1) a recognitional moral sensitivity enables people to notice ethical problems due to a refined understanding of ethical concepts and corresponding attitudes; (2) a sympathetic moral sensitivity enables people to notice ethical problems due to a concern for those who are affected by their actions; (3) a critical moral sensitivity enables people to notice when their sympathetic and recognitional moral sensitivity could be undermined or biased due to factors like conflicts of interest or fatigue. The second part of this dissertation investigates the nature of moral sensitivity from an empirical perspective. Studies were analyzed and synthesized, which can legitimately claim to investigate moral sensitivity based on valid tests of people’s ability to notice ethical problems. The findings support and help to refine the idea that moral sensitivity bears a recognitional, a sympathetic, and a critical component. Moral expertise, i.e., refined knowledge of ethical concepts, is shown to be indispensable for the recognition of subtle ethical problems – particularly when people have competing objectives or face stress. Moral expertise, on its own, appears to be insufficient, however, if it is not combined with appropriate attitudes towards ethical problems, which motivate people to notice them in the first place, and which trigger appropriate emotional responses in people (e.g., guilt at the thought of ignoring someone’s will). Contrarily, many social and emotional competences that have been viewed as helping people to notice ethical problems, e.g., perspective-taking skills, do not appear to be reliable sources of moral sensitivity. What seems to make a big difference, however, is whether people hold sympathetic and unprejudiced attitudes towards the people who are affected by their actions, as a lack of sympathy and demeaning attitudes will inhibit people’s ability to notice when their actions may be immoral. Finally, the evidence culminates in showing how moral sensitivity can be undermined and biased in various ways, e.g., due to stronger empathic reactions to people who are immediately present. To cope with various sources of bias, people also need to cultivate an automatized ability to recognize when their moral perception could be biased. The final third of this dissertation concentrates on the question, how digital games can be used to promote moral sensitivity. Building on the unified theory of moral sensitivity that was established in the prior parts, strategies were identified that are generally effective in promoting moral sensitivity. The most effective approach appears to be experiential learning in combination with feedback on people’s ability to notice ethical problems. Digital games are found to provide an optimal environment for the provision of this kind of training, because they can motivate learners to engage with ethical problems in an interactive manner, provide appropriate feedback, and do all of this in a safe context. Twenty-four game mechanisms are explored in view of their potential to promote recognitional, sympathetic, and critical moral sensitivity, whereby the findings of more than 80 empirical studies are synthesized. Several game mechanisms are found to be highly promising, while there is a lack of empirical information on others, and some mechanisms appear to bear risks, e.g., the use of egoistic temptations. A critical discussion of different ways in dealing with the risk of malicious play, i.e., the performance of unethical actions in games, suggests that digital games could undermine moral sensitivity, if they provide players with reasons to justify malicious play. On the other hand, when bad choices and malicious play are confronted with negative feedback and bad consequences, the possibility to make ethical mistakes appears to be legitimate and highly useful.