Ambiguity aversion is one of the most robust phenomena documented in the decision making literature, though there are still disagreements concerning what constitutes an adequate account for its occurrence. We describe six experiments that shed additional light on the enigmatic phenomenon of ambiguity avoidance. Experiments 1 and 2 demonstrate that ambiguity avoidance holds for both gains and losses and for different modes of framing. It holds even when the ambiguous condition is normatively superior. Experiments 3–6 vary the amount of information associated with ambiguity. We conclude that perceived informativeness (regardless of its normative relevance) is a major component in distinguishing between risk and ambiguity, as well as determining the level of ambiguity. In the final discussion we link ambiguity avoidance to the ‘certainty effect' (Kahneman & Tversky, 1979), and suggest that both should be treated in the same framework.