The number and type of scoring errors made on Holland's Self-Directed Search-R (SDS-R) are examined. participants were high school juniors and seniors. The results indicate that although some errors were made, the SDS-R was perceived as a useful career assessment tool. Implications of the results are also discussed.The Self-Directed Search (SDS; Holland, Powell, & Fritzsche, 1994) is a self-administered, self-scored, and self-interpreted career counseling tool. The SDS was developed for two primary reasons: (a) to multiply the number of people a counselor can serve, and (b) to provide a career counseling experience for people who do not have, or choose not to access, a vocational counselor. Even though there are more than 100 revisions on the newest versions of the SDS, the reliability and validity of any self-administered, self-scored instrument can be compromised through error in self-administration and self-scoring (Cummings & Maddux, 1987).For example, Gelso, Collins, Williams, and Sedlacek (1973) found that as many as 50% of incoming college freshman make serious scoring errors on the Self-Directed Search and that less than half feel that their results were reasonable. Other researchers (e.g.. CMstensen, Gelso, Williams, & Sedlacek. 1975; ONeil, Price, & Tracey, 1979) have found similar results using the SDS. In a study comparing error rates on the original SDS in 1970, and on the 1977 revision, Tracey and Sedlacek (1980) found that the 26% error rate affecting the summary code for the revised edition was significantly lower than the 54% error rate for the original SDS. No differences, however, were found between the original and revised forms in the error rate (18% and 12%. respectively) affecting high-point codes. Finally, in a study comparing the scoring errors of learning disabled and nonlearning disabled students on the SDS, Cummings