Cormorants are assumed to have a “partially wettable” plumage as a mechanism to reduce buoyancy while swimming underwater. This assumption is mainly based on 3 observations: 1) the volume of air in the plumage of submerged carcasses is small compared to other water birds, 2) cormorants assume a “wing drying” posture when they exit the water, and 3) the feather structure of the plumage. An alternative mechanism to reduce buoyancy is to release air out of the plumage by ptilomotion without allowing water to penetrate. How wet cormorants actually get is an open issue that has important implications for the energy budget of these warm blooded aquatic predators. Here we report empirical measurements on the amount of water retained in the plumage of live great cormorant Phlacrocorax carbo sinensis during voluntary swimming and diving in an experimental design that simulates a foraging diving bout. The amount of water retained in the plumage increased as a function of time spent in water. However the birds limited their dive bouts to less than 18 minutes so that the added mass of retained water did not exceed 6% of their body mass. This maximal level of water retention is estimated to reduce the buoyancy of the dry bird by 18%. This maximal level is also similar to measurements of water retention of carcasses and suggests that measurements preformed on carcasses yield only the upper level of water absorption while live birds slow down water penetration, allowing longer periods of foraging.