A study investigated the use of riddles beginning with How, When, and Why to see which question type was easier for low-achieving students to answer correctly, and to see if practice could increase the number of correct answers chosen. For comparison purposes, scores were also collected for an average-achieving group without practice, to see if practice could decrease the difference between low and normal achievers. Subjects for the practice group were 48 second and third graders at an inner-city school who were enrolled in a pull-out reading program for students reading below grade level. Subjects for the comparison group were 42 children in the same grades at the same school reading at grade level. Forty-eight riddles (16 of each type) were used. The riddles were read by the examiner to the practice group who would have had difficulty reading them alone. Comparison of pretest and posttest results indicate that How riddles were the easiest to answer for both groups, and When riddles were easier for both groups to answer correctly than Why riddles. The low achievers' response to the riddles was very positive, and the fact that this practice group performed as well as they did suggests that riddles may be a useful tool in refining language concepts in reading. (Tables of data and references are appended.) (NKA)