We examine the sensitivity of simultaneous simulation of climate, natural 14C, bomb 14C, and anthropogenic CO2 uptake to the choice of three different ocean mixing schemes: horizontal/vertical mixing (HV), isopycnal mixing (ISO), and Gent-McWilliams mixing (GM) using an Earth system model of intermediate complexity, Integrated Science Assessment Model-2.5D (ISAM-2.5D). Our modeling results suggest that the HV scheme greatly underestimates the observed values of natural 14C in the deep ocean, while the ISO and GM schemes yield more realistic results by simulating increased amounts of natural 14C values through enhanced vertical diffusion and deep water formation. The GM scheme further improves the ISO-based natural 14C distribution in the Southern Ocean and the deep Pacific and Indian oceans through a more realistic simulation of the Southern Ocean circulation. The model simulated global uptake of anthropogenic CO2 for the 1980s ranges between 1.8 and 2.3 PgC/yr, largely consistent with data-based estimates and OGCM results. The ISAM-2.5D simulated oceanic uptake of 14C and CO2 is highest for the ISO scheme and lowest for the HV scheme, with the largest discrepancies occurring among different mixing schemes found in the Southern Ocean. However, no single mixing scheme is more successful than the others in simulating GEOSECS-measured uptake of bomb 14C and anthropogenic CO2 for various ocean basins. Climate change is found to reduce CO2 uptake by 7--9% and 6--8% for the 1980s and over the period 1765--1990, mainly as a result of decreased CO2 solubility associated with increased sea surface temperatures. However, the effect of climate change on bomb 14C uptake is negligible.