In the case of applying shape memory alloy (SMA) to an actuator, the greatest problem is its nonlinear behavior. It is also known that this characteristic of SMA significantly depends on working conditions and environment. Consequently, development of a mathematical model of the SMA actuator is remarkably difficult. In this study, a method is investigated for realizing position control of the SMA actuator using fuzzy reasoning. In this reasoning, qualitative representation that is based on the rule of linguistic variables is easy to make. As a result, the identification of a mathematical model for control of the system becomes unnecessary. In this paper fuzzy reasoning is applied to control of the SMA actuator, and its applicability is confirmed through experiments.