This study examines the accuracy of the self-scoring efforts of 306 eighth-graders on the Kuder General Interest Survey (GIS), and suggests possible methods to improve self-scoring accuracy. The GIS is widely used to assist junior high school students with their educational and vocational planning. After the administration of the test by English teachers (not aware of the research), the students' self-scored sections were checked for accuracy by an independent counter. Analysis results indicated sizable student error (94 counting errors were of magnitude 10 or more) that could definitely change the shape of a student's interest profile. Factors contributing to scoring error were not isolated. Suggestions for survey administrators to reduce self-scoring error include: provide adequate time; carefully supervise the students' pin hole placements and scale use; have students recheck their counts; or use the machine scored version. (BS)