To study the relationship between human lifestyle and health, comprehensive data on individuals’ daily activities and health information are essential. Although several datasets containing lifestyle and health information exist, comprehensive datasets are rare. In our project, we have compiled and published such a comprehensive dataset. This dataset is a comprehensive collection of behavioral, biometric, and environmental data collected from 39 participants who spent between 1 week and 2 months in smart rooms in Tokyo, Japan, over a period of approximately 3 years. This includes personal data such as household appliance use, heart rate, sleep quality, temperature, lighting, and detailed meal data with associated images. While there are many datasets that publish these types of data individually, the unique value of our dataset lies in its comprehensive nature, linking all such data to individual IDs. Data were collected over 732 days, resulting in 107,266,968 records, including 900 meal images. The total size of the dataset was 4.30 GB. This dataset is invaluable for machine learning and analysis, providing insights for improving sleep quality, among other potential applications.