Two investigations tested the hypothesis that the need for a randomly determined outcome influences the degree of perceived skill involved in attaining the outcome, which, in turn, influences the confidence in attaining the outcome. In Study 1, food-deprived or food-satiated subjects expected to have a good or poor chance of winning a food incentive in a drawing. Compared to food-satiated subjects, food-deprived subjects thought the drawing involved more skill, and they were more confident about winning. In Study 2, state lottery players reported their incomes, how much skill they thought playing the lottery involved, and how confident they were about winning Path analyses showed that income negatively affected skill ratings, and skill ratings positively affected confidence-in-winning ratings. Results were replicated using comparable questions about a lottery-sponsored rub-off card game. Results are discussed in terms of Langer's propositions regarding the illusion of control.