This dissertation focusses on the question how social mechanisms, such as social contagion, communication, and group formation, affect blood donation behaviour. Blood donations enable transfusions, many routine medical treatments, and the production of plasma-derived medicinal products. At the same time, they are a textbook example of real-world cooperation, where individuals make costly contributions to a public good. In four empirical chapters studying blood donation in the Netherlands and Australia, this thesis shows that a) heterogeneity in blood donor motivation can be mapped into four donor types, which are associated with blood donor background characteristics and long-term donor lapse, b) there is social contagion in blood donation, such that blood donors increase or decrease their donations in line with the donations made by other donors within their neighbourhood and especially their partners, c) talking about donations but not recruitment via word of mouth predicts compliance with solicitations for donations that are sent out by the blood bank, and d) the formation of identity-based groups among donors is highly effective for increasing donation frequency. The main contribution of this thesis to the literature is to link the largely individual-centred blood donation literature with the experimental literature on social mechanisms that sustain cooperation, and to empirically demonstrate the benefits of this perspective for understanding blood donation behaviour and blood banking practice.