The higher level processes such as monitoring and controlling cognitive behavior have been pervasively emphasized and metacognitive regulation has been regarded as one of the hot-points of contemporary psychological studies in recent years. Metacognitive regulation is generally recognized to be an advance cognitive mechanism, which exert huge impact on cognitive effectiveness. Researches on metacognitive regulation have important implications for theoretical development and practice. Metacognitive regulation is such a supreme ability that researchers want to distinguish individuals by their regulation levels. However, its conceptual structure is still under discussion. Nelson and Narens suggested that metacognitive regulation was composed of monitoring and control processes interactively, while Yeager split metacognitive regulation into another two parts, named (a) Prediction and Planning, and (b) Study Techniques and Their Control. Both theories sound reasonable, but not convincible. According to the object, area and time of regulation, we decomposed metacognitive regulation into two components: on-line regulation and off-line regulation. On-line regulation happens in the performance of a certain task, including monitoring and control of the problem solving processes. Whereas off-line regulation takes place out of a certain task situation, reflecting individual’s self-regulation, planning and re-evaluation of his behaviors on a global level. The purpose of the present study was to explore the psychometric structure of off-line metacognitive regulation, and to develop an assessing instrument with high criterion-related validity. The data of this study were collected from a large-scale IT company in Beijing. In study 1, 226 employees participated the preliminary test. By using a self-developed questionnaire, a three-factor structure of off-line metacognitive regulation was obtained, which included global-planning, awareness and generalization. After a retest study with 115 participants, the structure of off-line metacognitive regulation was confirmed. In study 2, hierarchical linear regression and hierarchical logistic regression were conducted to examine the criterion-related validity of the off-line regulation questionnaire. The results showed that managers were different from common staffs in particular regulation factors. Managers have higher generalizing level than others in sales departments, while managers in support department have higher awareness level than others. But on the other hand, for the common staffs, the level of off-line metacognitive regulation failed to predict employees’ performance in sales department. However, for the staffs in support department, the level of global-planning predicted the creative activities reversely. Based on the results, we may suggest that the off-line metacognitive regulation questionnaire has good psychometric quality, and the procedures of administrating and scoring are objective and standardized. It can be applied to practice such as personnel selection and promotion.