A pneumatic servo seems to be effectively applied to many varieties of gripping or compressing works,because its high compressibility is useful to the force control.In the force control system,the characteristic of the object is comprised in the closed loop,so its influence has to be considered in the design of controller.However,it is generally unknown and often changes during operation.In particular,the characteristics of the plant changes significantly,when the object is moving.To cope with these problems,more intelligent control schemes are required.In this paper,we propose a design scheme of a model reference adaptive control with Neural Network for a force control system with pneumatic servo.The effectiveness of the proposed design scheme is confirmed by experiments using the existent pneumatic servo system..