MIMIC-IV is a publicly available electronic health record (EHR) dataset sourced from the Beth Israel Deaconess Medical Center, covering admissions between 2008 and 2019. The dataset includes patient measurements, orders, diagnoses, procedures, treatments, and deidentified free-text clinical notes. It aims to support a wide array of research studies and educational materials, reducing barriers to conducting clinical research. MIMIC-IV is structured with a modular organization, allowing linkage to external departments and different modalities of data. The dataset is accessible via PhysioNet, with the hosp and icu modules available on the MIMIC-IV project page, and the note module on the MIMIC-IV-Note project page. The creation of MIMIC-IV involved data acquisition, transformation, and deidentification processes, ensuring data integrity, consistency, and deidentification. Users must complete a training course and sign a data use agreement to access the dataset. The MIMIC Code Repository provides scripts for extracting subsets of information and supports cloud-based interrogation of the dataset.MIMIC-IV is a publicly available electronic health record (EHR) dataset sourced from the Beth Israel Deaconess Medical Center, covering admissions between 2008 and 2019. The dataset includes patient measurements, orders, diagnoses, procedures, treatments, and deidentified free-text clinical notes. It aims to support a wide array of research studies and educational materials, reducing barriers to conducting clinical research. MIMIC-IV is structured with a modular organization, allowing linkage to external departments and different modalities of data. The dataset is accessible via PhysioNet, with the hosp and icu modules available on the MIMIC-IV project page, and the note module on the MIMIC-IV-Note project page. The creation of MIMIC-IV involved data acquisition, transformation, and deidentification processes, ensuring data integrity, consistency, and deidentification. Users must complete a training course and sign a data use agreement to access the dataset. The MIMIC Code Repository provides scripts for extracting subsets of information and supports cloud-based interrogation of the dataset.