Objective To determine the prevalence of clinically relevant drug-drug interactions in an intensive care unit of a tertiary care hospital in the United States and to compare to an intensive care unit at a Pakistani hospital, which lacks electronic medical record-based drug-drug interaction screening. Study setting A retrospective cross-sectional analysis was conducted in the cardiovascular intensive care unit (CVICU) at Michigan Medicine (MM), Ann Arbor, MI, USA between Jan 2018 – Jan 2019. Study Design Analysis of 300 MM patients was conducted to identify drug-drug interactions using Micromedex® and Lexicomp®. Descriptive statistics and multivariate binary logistic regression was used. Independent samples t-test was used to compare prevalence between MM and in a similar cohort of patients in the cardiac intensive care (CCU) at KTH, Pakistan from a previously published study. Data Collection Data was collected for patients who were admitted to the CVICU for at least 24 hours and were prescribed at least 2 drugs from the electronic health record of MM. Principal Findings In the intensive care unit of the US hospital, 58% of patients had at least one drug-drug interaction, while 16% had a clinically relevant drug-drug interaction. Significantly fewer patients had drug-drug interactions at the US hospital than the Pakistani hospital (58% vs. 95%, p < 0.01). Polypharmacy and length of stay increased drug-drug interaction occurrence in the US hospital (p <0.01). Conclusion The prevalence of drug-drug interactions in the intensive care unit at the US hospital was high but lower than the Pakistani hospital, likely due to electronic medical record-based screening. Despite electronic medical record-based screening at the US hospital, 8 clinically relevant drug-drug interaction pairs were undetected.