A study was undertaken to determine the operational reading levels and skills of beginning office employees and to compare the readability of classroom and on-the-job materials. Clerks and secretaries who had been employed for two years or less were observed and interviewed to collect the data. Statistical analysis revealed that secretaries read significantly different materials than did general clerks. Most of the reading was for verifying, comprehending, and proofreading; but the secretaries verified different materials than clerks, while the general office clerks read more materials for comprehension. Secretaries read more rough draft materials, whereas clerks read more printed materials. There was an approximate balance in the reading of typed materials between the two groups. Readability of on-the-job materials varied between 10.5 for order forms/invoices to 12.9 for labels on packages. No significant readability differences were found between instructional and on-the-job materials. It appeared that textbooks, an instructional simulation, and on-the-job materials provided similar distribution of comprehending and proofreading materials, though simulation materials emphasized more verifying skills than on-the-job materials. (RL)