A group of inadequate readers was compared with adequate readers on measures of saccadic eye movements. Saccades are discussed as a possible contributing cause of reading problems rather than the traditional view that they are primarily or solely the result of reading experience. Nonreading as well as reading materials and tasks were used. On all nonreading and reading materials the inadequate readers tended to make more regressions and fewer forward fixations. The inadequate readers were heterogeneous; two subgroups involving micro-sequencing evolved. Implications and procedures for saccadic diagnosis are discussed.