Fifty-nine elementary school children (29 normally achieving and 30 reading-disabled) without uncorrected refractive error or ocular pathology were tested with the Developmental Eye Movement Test (DEM). They were retested a month later. For the reading disabled group, there were no significant test-retest differences in the vertical reading time, horizontal adjusted reading time, ratio of horizontal to vertical time, or the number of errors. For the normally-achieving group, there was only a significant difference in the vertical reading time. Overall test-retest correlations were moderate to high (0.85 for errors, 0.86 for vertical time, 0.79 for horizontal time, and 0.52 for ratio). Ten children did not follow any of the four expected oculomotor types. On those that followed, test-retest correlation was moderate (0.59). The DEM promises to be a reliable clinical tool in oculomotor assessment, especially by further refinement of the classification of oculomotor types.