Although the diagnosis of strabismus requires specialist examination, many individuals are aware that they are affected. It is thus possible that self-reporting could be sufficient for population or genetic studies of strabismus; however, the accuracy of self-reported strabismus has not previously been evaluated. In this study, participants in the Genetics of Comitant Congenital Strabismus (CCS) Study were asked to report whether they had strabismus prior to receiving a complete orthoptic evaluation. In 671 individuals studied, the sensitivity of self-report for detecting true CCS was 48.5%, with a specificity of 98.6%, giving a PPV of 92.6% (NPV 84.5%). Self-reporting accuracy was influenced by the direction, size and constancy of the deviation, and by sex but not education. Self-reports produced a misclassification rate of 14.5% for CCS alone and 33.1% for combined CCS or strabismus-associated conditions. Considering this high misclassification rate, self-report should not be used for clinical studies of strabismus.