A mapping procedure is presented for studying the construct validity of interest inventories that assess the six interest types defined by J. L. Holland. The 1989 revision of the American College Testing Program Interest Inventory (UNIACT) was used to assess the interests of 1,078 12th-grade students (497 males and 581 females) and 725 adults aged 25 years or older (287 males and 438 females). The revised UNIACT contains six 15-item scales: business contact, business operations, technical, science, arts, and social service. Items cover work-relevant activities via a three-choice response format consisting of "dislike", "indifferent", and "like." Respondents were allocated to career groups on the basis of career choice (12th-graders) or current occupation (adults). The interests of 27 career groups were mapped onto Holland's hexagon using scores for the underlying data/ideas and things/people dimensions. Expected group locations were determined via expert judgment for 11 of these groups. Mean interest scores were also profiled. Results from both samples support the UNIACT's construct validity. Career group locations on the hexagon generally agreed with experts' judgments and common sense. The advantages of the mapping procedure over multiple-score profiles are considered. Four tables present study data and nine figures illustrate the relationships. Five appendices present a World-of-Work Map, worksheets used by the expert panel, occupational group classifications for 12th-graders and adults, and criterion group hit rates based on high point codes. A 24-item list of references is included. (Author/SLD)