Most organizations planning to assess adverse impact or perform a stock analysis for affirmative action planning must correctly classify their jobs into appropriate occupational categories. Two methods of job classification were assessed in a combination archival and field study. Classification results from expert judgment of functional job analyses were compared to statistical profiles of the Position Analysis Questionnaire (PAQ). From the data banks of the PAQ services, 300 cases were randomly selected. In order to obtain the Equal Employment Opportunity (EEO) classification for the selected cases, the jobs were first categorized by industry. Major corporations in each of the industrial areas were selected, and personnel officers of the selected corporations were sent a questionnaire containing the Dictionary of Occupational Titles (DOT) written job descriptions. They were asked to place the jobs into one of the three EEO classifications. A discriminant analysis was performed and statistical results slightly favored the PAQ. The PAQ classification was 72 percent correct compared to 70 percent for functional job analysis. Results strongly support the notion that quantitative job analysis data can be successfully used to classify jobs into their EEO categories. The primary difference in the two systems is in their ability to correctly classify unskilled jobs, where the PAW job dimensions do considerably better than the DOT worker function scales. (JAC)