To John Rawls, the basic structure of society is the determining factor in the distribution of benefits and burdens, and as such, the primary subject of justice. If citizens were to choose the principles governing this structure – behind Rawls’ veil of ignorance, the original position – Rawls would assume them to choose along the maximin decision rule; maximising the worst possible outcome. John Harsanyi rejects Rawls’ assumption of choosing in accordance with maximin, as to him, this would suggest irrationally high levels of risk-aversion by decision makers. In this paper, I will argue against Harsanyi’s criticism of the maximin principle, mainly by questioning the validity of (subjective) probability calculations in the original position. By further pointing towards a key distinction between risk-aversion and uncertainty-aversion, I suggest that Harsanyi’s premise of extreme risk-aversion loses its force. From there, I conclude that Rawls’ inclination towards the maximin principle in establishing his two principles of justice appears defensible.