Reconciliation processes have attracted much scientific interest over the last two decades. In the present paper, after discussing the difficulty in defining the elusive concept of reconciliation, I will present the theoretical perspective of the Needs-based Model of reconciliation. According to this model, transgressions threaten the identities of victims and perpetrators in an asymmetrical manner. Victims experience threat to their sense of agency, whereas perpetrators experience threat to their moral image. The restoration of victims’ and perpetrators’ positive identities (e.g., through the exchange of empowering and accepting messages, or through identity-affirmation interventions) may increase their willingness to reconcile with each other. I will present empirical evidence supporting the Needs-based Model’s hypotheses in various contexts including interpersonal transgressions, direct intergroup violence, structural inequality, and duality of social roles (i.e., when both parties transgress against each other and engage in “competitive victimhood”). Finally, I will discuss the practical implications of the model and point to future research avenues.