This study observed human motions of whipping conducted by five expert confectionery hygiene mistresses, because mixing performance can be easily estimated by change of physical properties in heavy cream. Mixing state of heavy cream was evaluated by an overrun which indicates the amount of gas phase in the cream. Rheological properties of the cream were also measured for mixing state estimation. Agitating behaviors were recorded by a high-speed camera. Fluid flow motion was numerically observed by the Moving Particle Semi-implicit (MPS) method. The experts' motions were classified into two patterns, i.e. vertically elliptical rotation and horizontally reciprocating. Numerical analysis suggested that these motions with large amplitude produced large disturbances and high shear on the liquid surface and consequently enhanced the uptake of air as fine bubbles. It is considered that elliptical and reciprocating motions are effective for whipping because the shearing on fluid and efficient air uptake improve the whipping performance.