Eyewitnesses play an important role in the justice system. But suggestive questioning can distort eyewitness memory and confidence, and these distorted beliefs influence jurors (Loftus, Learning & Memory, 12, 361-366, 2005; Penrod & Culter, Psychology, Public Policy, and Law, 1, 817-845, 1995). Recent research, however, hints that suggestion is not necessary: Simply changing the order of a set of trivia questions altered people's beliefs about their accuracy on those questions (Weinstein & Roediger, Memory & Cognition, 38, 366-376, 2010, Memory & Cognition, 40, 727-735, 2012. We wondered to what degree eyewitnesses' beliefs-and in turn the jurors who evaluate them-would be affected by this simple change to the order in which they answer questions. Across six experiments, we show that the order of questions matters. Eyewitnesses reported higher accuracy and were more confident about their memory when questions seemed initially easy, than when they seemed initially difficult. Moreover, jurors' beliefs about eyewitnesses closely matched those of the eyewitnesses themselves. These findings have implications for eyewitness metacognition and for eyewitness questioning procedures.