MIMIC-III is a large, freely accessible critical care database containing information on patients admitted to critical care units at a large tertiary care hospital. It includes vital signs, medications, laboratory measurements, observations, and notes charted by care providers, fluid balance, procedure codes, diagnostic codes, imaging reports, hospital length of stay, survival data, and more. The database supports academic and industrial research, quality improvement initiatives, and higher education coursework. MIMIC-III was developed to address the challenges of data integration in digital health systems and to improve the reproducibility of clinical research. It is an update to the widely-used MIMIC-II database and includes deidentified, comprehensive clinical data of patients admitted to Beth Israel Deaconess Medical Center in Boston, Massachusetts. The open nature of the data allows clinical studies to be reproduced and improved. MIMIC-III contains data associated with 53,423 distinct hospital admissions for adult patients (aged 16 years or above) admitted to critical care units between 2001 and 2012, and data for 7870 neonates admitted between 2001 and 2008. The median age of adult patients is 65.8 years, 55.9% are male, and in-hospital mortality is 11.5%. The median length of an ICU stay is 2.1 days, and the median length of a hospital stay is 6.9 days. A mean of 4579 charted observations and 380 laboratory measurements are available for each hospital admission. The database includes data from two critical care information systems: Philips CareVue Clinical Information System and iMDsoft MetaVision ICU. Data was merged when building the database tables, with exceptions for data relating to fluid intake. The data was deidentified in accordance with HIPAA standards using structured data cleansing and date shifting. The MIMIC-III database is a relational database consisting of 26 tables. It includes data on patient stays, dictionaries for cross-referencing codes, and data associated with patient care. The database is available for research purposes, with researchers required to formally request access via a process documented on the MIMIC website. The database is also used for coursework and research studies, including machine learning approaches for predicting patient outcomes, clinical implications of blood pressure monitoring techniques, and semantic analysis of unstructured patient notes. The MIMIC-III database is supported by a public code repository, and efforts are underway to transform MIMIC into common data models such as the Observational Medical Outcomes Partnership Common Data Model. The database is licensed under a Creative Commons Attribution 4.0 International License.MIMIC-III is a large, freely accessible critical care database containing information on patients admitted to critical care units at a large tertiary care hospital. It includes vital signs, medications, laboratory measurements, observations, and notes charted by care providers, fluid balance, procedure codes, diagnostic codes, imaging reports, hospital length of stay, survival data, and more. The database supports academic and industrial research, quality improvement initiatives, and higher education coursework. MIMIC-III was developed to address the challenges of data integration in digital health systems and to improve the reproducibility of clinical research. It is an update to the widely-used MIMIC-II database and includes deidentified, comprehensive clinical data of patients admitted to Beth Israel Deaconess Medical Center in Boston, Massachusetts. The open nature of the data allows clinical studies to be reproduced and improved. MIMIC-III contains data associated with 53,423 distinct hospital admissions for adult patients (aged 16 years or above) admitted to critical care units between 2001 and 2012, and data for 7870 neonates admitted between 2001 and 2008. The median age of adult patients is 65.8 years, 55.9% are male, and in-hospital mortality is 11.5%. The median length of an ICU stay is 2.1 days, and the median length of a hospital stay is 6.9 days. A mean of 4579 charted observations and 380 laboratory measurements are available for each hospital admission. The database includes data from two critical care information systems: Philips CareVue Clinical Information System and iMDsoft MetaVision ICU. Data was merged when building the database tables, with exceptions for data relating to fluid intake. The data was deidentified in accordance with HIPAA standards using structured data cleansing and date shifting. The MIMIC-III database is a relational database consisting of 26 tables. It includes data on patient stays, dictionaries for cross-referencing codes, and data associated with patient care. The database is available for research purposes, with researchers required to formally request access via a process documented on the MIMIC website. The database is also used for coursework and research studies, including machine learning approaches for predicting patient outcomes, clinical implications of blood pressure monitoring techniques, and semantic analysis of unstructured patient notes. The MIMIC-III database is supported by a public code repository, and efforts are underway to transform MIMIC into common data models such as the Observational Medical Outcomes Partnership Common Data Model. The database is licensed under a Creative Commons Attribution 4.0 International License.