Improvement of mental health of workers has recently attracted much attention because it is strongly associated with the reduction of healthcare expenses. The critical changes in mental conditions leading to the development of diseases, such as psychiatric disorders, would be signaled by alterations in psycho-physiological states or habitual behaviors in daily life. Therefore, their accurate evaluation and continuous monitoring might be essential for the management of mental health. Meanwhile, a variety of factors, such as labor conditions or environments, personal relations at work, affect mental conditions in workers. Thus, the integrated monitoring and analysis of these information is also required. In this study, we used a smartphone data collector capable of Ecological Momentary Assessment, which is recognized as a gold standard method with having a higher ecological validity, to record subjective symptoms of workers in daily life. In addition, behavioral and physiological information of workers and their surrounding environmental information were simultaneously measured (about 200 workers; about 1 week). We show the utility of multilevel modeling approaches by showing some examples analyzed by using the large-scale hierarchical data set collected in this study.