Abstract Historically the concept of risk is rooted in Renaissance lifestyles, in which autonomous agents such as sailors, warriors, and tradesmen ventured upon dangerous enterprises. Thus, the concept of risk inseparably combines objective reality (nature) and social construction (culture): Risk = Danger + Venture. Mathematical probability theory was constructed in this social climate in order to provide a quantitative risk assessment in the face of indeterminate futures. Thus we have the famous formula: Risk = Probability (of events) × the Size (of future harms). Because the concept of harm is always observer relative, however, risk assessment cannot be purely quantitative. This leads to the question, What are the general conditions under which risks can be accepted? There is, after all, a difference between incurring a risk and bearing the costs of risks selected for by other agencies. Against this background, contours of a theology of risk emerge. If God creates a self-organizing world of relatively autonomous agents, and if self-organization is favored by cooperative networks of autopoietic processes, then the theological hypothesis of a risk-taking God is at least initially plausible. Moreover, according to the Christian idea of incarnation, God is not only taking a risk but is also bearing the risks implied by the openness of creation. I thus argue for a twofold divine kenosis—in creation as well as in redemption. I discuss some objections to this view, including the serious counterargument that risk taking on behalf of others remains, even for God, a morally dubious task. What are the conditions under which the notion of a risk-taking God can be affirmed without leaving us with the picture of God as an arbitrary, cosmic tyrant? And what are the practical implications for the ways in which human agents of faith, hope, and love can learn to cope with the risks of everyday life and of political decisions?