This paper deals with the control of hand transfer trajectory using fuzzy reasoning. The motion, when an object to transferred from one point to another in the two-dimensional plane, is classified into five modes. These trajectories of five modes are shown by fourth linear differential equations which are derived through the optimal control theory. Firstly, fuzzy control rules selecting the position in pre-condition and the velocity in conclusion are proposed based on those trajectories. It is shown that the trajectories in all five modes are well reproduced by the proposed rules. Secondly, by taking the differential of acceleration around the start point into consideration, it is shown that smooth trajectories like a human motion are simulated.