In a laboratory-based test series, seven experiments along a simulated Pacific hydrotherm at 152 degrees W, 40 degrees N were carried out to measure the rise velocities of liquefied CO2 droplets under (clathrate) hydrate forming conditions. The impact of a hydrate skin on the rising behavior was investigated by comparing the results with those from outside the field of hydrate stability at matching buoyancy. A thermostatted high-pressure tank was used to establish conditions along the natural oceanic hydrotherm. Under P-/T-conditions allowing hydrate formation, the majority of the droplets quickly developed a skin of CO2 hydrate upon contact with seawater. Rise rates of these droplets support the parametrization by Chen et al. (Tellus 2003, 55B, 723-730), which is based on empirical equations developed to match momentum of hydrate covered, deformed droplets. Our data do not support other parametrizations recently suggested in the literature. In the experiments from 5.7 MPa, 4.8 oC to 11.9 MPa, 2.8 degrees C positive and negative deviations from predicted rise rates occurred, which we propose were caused by lacking hydrate formation and reflect intact droplet surface mobility and droplet shape oscillations, respectively. This interpretation is supported by rise rates measured at P-/T-conditions outside the hydrate stability field atthe same liquid CO2-seawater density difference (delta rho) matching the rise rates of the deviating data within the stability field. The results also show that droplets without a hydrate skin ascend up to 50% faster than equally buoyant droplets with a hydrate skin. This feature has a significant impact on the vertical pattern of dissolution of liquid CO2 released into the ocean. The experiments and data presented considerably reduce the uncertainty of the parametrization of CO2 droplet rise velocity, which in the past emerged partly from their scarcity and contradictions in constraints of earlier experiments.