Project Abstract Substance use disorder (SUD) diagnosis and treatment requires laboratory drug testing, which inflicts monetary, time, and interpersonal barriers to professional care. To date, there are no remote, non-invasive and instant opioid testing options available, limiting the potential for clinical outreach. Tenvos’ solution is an Artificial Intelligence based technology that detects opioid use from 15 seconds of speech. Tenvos allows clinicians to screen for opioid use remotely, mitigating the barriers to clinical intervention, streamlining the clinical workflows, reducing costs, and improving the overall quality of patient care. Tenvos’ solution is an easily-deployable API-first product, allowing for seamless integration into various workflows– integration with virtual health platforms, phone appointments, or large scale automatic phone-based screening campaigns. This new approach will increase access for rural and low-income populations as well as incarcerated individuals. This project aims to further develop Tenvos’ solution and the understanding of opioids’ effects on voice production, while comparing the detection solution to current gold standard tests. The specific aims for the project’s Phase I set of activities include: Aim 1: Collecting clinical data from patients before and after the opioid treatment. Aim 2: Identifying 10 vocal characteristics that are affected by opioid use and ranked from Most Affected to Least Affected. Aim 3: Determining the accuracy of the machine learning model trained on this data measured by Sensitivity and Specificity and 95% CI. This project will achieve the aforementioned aims through a multisite observational study at Doctors Medical Center and Loma Linda University Health. Selected patients in the sites’ emergency departments will provide voice samples prior to and following opioid administration. This data collection will allow for vocal analysis to identify characteristic differences between the voice samples taken before opioid administration and those taken after. Furthermore, this sample collection will provide the initial feasibility data to facilitate the necessary steps toward commercialization.