Post-stroke delirium (PSD) is an important complication to identify in the inpatient rehabilitation setting due to its negative impact on rehabilitation outcomes. Delirium diagnosis is challenging and is rendered more so in the presence of post-stroke language and cognitive impairments. As delirium is a clinical diagnosis, the majority of PSD studies use a Diagnostic and Statistical Manual (DSM)–based “expert clinical diagnosis” as the reference standard. However, a diagnostic protocol is rarely defined, which undermines transparency, interrater reliability, and reproducibility in delirium research. This report describes the clinical diagnostic protocol for the “Post-Acute Stroke Delirium Assessment” study, a quality improvement initiative to be conducted in adult stroke rehabilitation inpatients at a tertiary academic hospital. The main study goals are to explore how delirium features present in this patient group and to use this information to create a new screening tool specific for PSD in inpatient rehabilitation. This protocol evaluates 16 delirium features, including those encompassed by the DSM, fifth edition, text revision as well as other features absent from the DSM, and delineates how to determine whether each feature is “present, absent, or unable to assess.” This protocol may also be valuable in informing delirium diagnosis in other patient groups with or without neurological deficits.