Background: Positive end-expiratory pressure (PEEP) is used to optimize oxygenation by preventing alveolar collapse. However, PEEP can potentially decrease cardiac output through cardiopulmonary interactions. The effect of PEEP on cardiac output during cardiopulmonary resuscitation (CPR) is not known. Methods: This was a preclinical randomized, controlled, animal study conducted in an animal research facility on 25 Landrace-Yorkshire pigs. After inducing cardiac arrest, CPR was performed with LUCAS 3. During CPR, pigs were ventilated at a PEEP of 0, 5, 10, 15, 20 cmH 2 O (randomly determined via lottery) for 9 min. Cardiac output, obtained via ultrasound dilution, and PaO 2 were measured, and oxygen delivery calculated for each PEEP. Results: A mixed-effects repeated-measures analysis of variance was used to compare the baseline value adjusted mean cardiac output, PaO 2 , and oxygen delivery between PEEP groups. Least significant difference test was used to conduct pairwise comparisons between PEEP groups. To determine optimum PEEP, Gaussian mixture model was applied to the adjusted means of cardiac output and oxygen delivery. Increasing PEEP to 10 and higher resulted in significant declines in cardiac output. A PEEP of 15 and higher resulted in significant declines in oxygen delivery. As PEEP was increased from 0 to 20, PaO 2 increased significantly. Gaussian mixture model identified the 0-5 PEEP group as providing optimal cardiac output and oxygen delivery, with PEEP of 5 providing the highest oxygen delivery. Conclusions: A PEEP of 0-5 resulted in the optimal oxygen delivery and cardiac output during CPR, with PEEP of 5 resulting in higher oxygen delivery, and a slightly lower, statistically insignificant cardiac output than PEEP of 0.