Diagnostic errors can lead to delayed or incorrect treatments, resulting in serious patient harms. Many patients seek second opinions in an attempt to mitigate this problem; however, the impact of these patient-initiated second opinions on outcomes has not been well defined. This study examined data from a large nationally administered program that allows patients to request second opinions from expert specialists. Second opinions led to a change in diagnosis or treatment in more than 40% of participants. The second opinion was judged to have moderate or major clinical impact on patients' diagnoses in approximately 21% of cases and on treatments in nearly 31%. It is not clear how often the second opinions were correct and whether they actually led to better patient outcomes. Even though 95% of participants were satisfied with the second opinion experience, only 61% planned to follow the expert's recommendation. As diagnostic errors garner more attention in patient safety, evaluating second opinion programs may help reveal patterns for identifying these types of errors.