The Ellsberg Paradox suggests that rational agents of the Bayesian kind do not ignore the quality of information at hand when making decision, as instead postulated by Leonard Savage in his foundational study on decision-making and later taken for granted in mainstream decision theory. In order to prove his argument Daniel Ellsberg proposed to represent the risk/uncertainty dichotomy in an experimental setting that evidenced the ambiguity of the decision environment, concentrating on choices related to urns containing coloured balls. This paper examines Daniel Ellsberg’s rationale underlying the paradox and his use of the urns, by placing his contribution in the context of the crucial biographical episodes of the 1950s and early 1960s that originated his interest in the paradox. It is argued that Ellsberg’s choice to use urn examples cannot be taken as indication of his focus on ambiguity and lack of interest in uncertainty proper. Rather, the issue of uncertainty was the main thread of his commitment to decision-making as field of applied research.