The reliability of estimations of working memory capacity has not been thoroughly examined. The present study examined the test-retest reliability for working memory capacity as estimated in a lateralized change detection task, which is frequently used in studies involving electroencephalography. The test-retest correlations between K values for each set size in the two tests varied from 0.502 to 0.757, with test-retest correlations rising as set size increased. The results indicate that individual visual working memory capacity can be reliably estimated in a change detection task. Furthermore, test-retest reliability was higher when the two tests occurred at the same time of day than at different times of day.