Milling head is an essential assembly in the five-axis computer numerical control machine tools, positioning precision of which directly affects the machining accuracy and surface quality of the processed parts. Considering the influence of nonlinear friction in the transmission mechanism and the uncertain cutting force disturbance on the control precision of the milling head, the static and dynamic performances of the milling head are analyzed; relationships among the drive torque, load torque, motion direction and system parameters are discussed; and, finally, nonlinear dynamic model of the milling head is established. A novel adaptive sliding mode control scheme based on the variable switching gain and the adjustable boundary thickness is proposed for this nonlinear dynamic model; the stability of the closed-loop system is guaranteed by the Lyapunov theory. Experimental results show that the proposed adaptive sliding mode control can reduce the chattering in the traditional sliding mode control and can achieve high control precision without knowing the boundaries of uncertainties in advance.