This article outlines a technique for selecting poor spellers, describes the type of remedial instruction administered in Buffalo Lake, Minnesota, in an effort to improve spelling ability, and summarizes the results of such instruction as measured by certain tests. In September, 1926, the spelling ability of each pupil in Grades III-VIII, inclusive, was measured by three different methods: i. A standard test. Each grade was given a test consisting of twenty-five words selected at random from the column of the Ayres Spelling Scale for which the standard percentage of accuracy for that grade is 84. Thus, Grade VI was given twenty-five words from Column M; Grade V, twenty-five words from Column 0; etc. The pupil's score on this test gave his spelling ability a rank as compared with an established norm.