The effect of half cycle slug size (HCSS) on Water Alternating Gas (WAG) performance in tertiary miscible carbon dioxide (CO 2 ) flooding is experimentally investigated. Different half cycle slug size influences the fluids mobility and the contact time between the injected fluids and oil, and hence to achieve the best possible displacement efficiency, an optimum water/gas half cycle slug size could exist and deserves investigation.Coreflood experiments are performed in Berea sandstone core, from which the WAG performance, such as percent oil recovery, tertiary recovery factor, and CO 2 utilization factor are determined. The core flooding experiments are conducted at 60 and at miscible condition, i.e., at a pressure 20% above the minimum miscible pressure (MMP) of the oil sample. The experiments utilize Cottonwood Creek crude oil and artificial brines. The injection and connate brine contain 33.33 wt% CaCl 2 and 66.67 wt% NaCl with salinities of 16000 ppm (mg/L) and 30000 ppm, respectively. In every core flood test, alternate cycles of brine and CO 2 with a WAG ratio of 1:1 are injected with half cycle slug size ranging from 0.05 to 0.75 PV.The results show that for the system of interest, the optimum HCSS is 0.1 PV, in which the CO 2 usage is only 0.6 PV to reach a high tertiary oil recovery of 40.63%. An HCSS higher or lower than 0.1 PV gives lower oil recovery. When HCSS is lower than 0.1 PV, the system gives lower oil recovery because some of the gas is trapped by water and keeps staying in the core instead of displacing oil. Meanwhile, an HCSS higher than 0.1 PV results in larger clusters of CO 2 and water, thus water inefficiently controls CO 2 mobility, which makes the CO 2 -oil contact time and interfacial area for mass transfer decrease and gas breakthrough occurs prematurely.This experimental study is an essential effort to obtain better understanding the effect of half cycle slug size on WAG performance, which has never been experimentally investigated before. The understanding is critical for optimizing the WAG performance.