Introduction: Cardiopulmonary arrests (CPAs) are common in the intensive care unit (ICU). However, effects of protocol deviations on CPA outcomes in the ICU are relatively unknown. Objectives: To establish the frequency of errors of commission (EOCs) during CPAs in the ICU and their relationship with CPA outcomes. Methods: Retrospective analysis of data entered into institutional registry with inclusion criteria of age >18 years and non-traumatic cardiac arrest in the ICU. EOCs consist of administration of drugs or procedures performed during a CPA that are not recommended by ACLS guidelines. Primary outcome: relationship of EOCs with likelihood of return of spontaneous circulation (ROSC). Secondary outcomes: relationship of specific EOCs to ROSC and relationship of EOCs and CPA length on ROSC. Results: Among 120 CPAs studied, there was a cumulative ROSC rate of 66%. Cumulatively, EOCs were associated with a decreased likelihood of ROSC (OR: 0.534, 95% CI: 0.387-0.644). Specifically, administration of sodium bicarbonate (OR: 0.233, 95% CI: 0.084-0.644) and calcium chloride (OR: 0.278, 95% CI: 0.098-0.790) were the EOCs that significantly reduced likelihood of attaining ROSC. Each 5-minute increment in CPA duration and/or increase in number of EOCs corresponded to fewer patients sustaining ROSC. Conclusions: EOCs during CPAs in the ICU were common. Among all EOCs studied, sodium bicarbonate and calcium chloride seemed to have the greatest association with decreased likelihood of attaining ROSC. Number of EOCs and CPA duration both seemed to have an inversely proportional relationship with the likelihood of attaining and sustaining ROSC. EOCs represent potentially modifiable human factors during a CPA through resources such as life safety nurses.