Scanning probe microscope (SPM) based nanomanipulations have been successfully applied to various fields to explore and study the unique structures and properties in the nano world. SPM has the ability to manipulate nanoparticles and modify a sample surface in nano scale. Therefore, the positioning accuracy of the SPM probe is the crucial for nanomanipulations. In general, the precision of SPM based nanomanipulations has been limited mainly by hysteresis, creep and drift of the piezo actuator. In this research, a new control strategy named hysteresis creep inverse based robust adaptive model reference control (RAMRC) is proposed to reduce the hysteresis, creep and system drift of the SPM scanner for improving the positioning accuracy of nanomanipulations. The RAMRC approach uses a compensator to compensate the hysteresis and creep. In addition, to tackle the unknown drift, the RAMRC also has a robust adaptive controller to explicitly deals with the noise and disturbance. This is suitable for manipulating nano-objects in a noisy environment, which is difficult for fixed operators/parameters based compensators and controllers. Additionally, the numerical experimental results support the RAMRC theory well and the position error can be controlled within approximate ten nanometers based on simulation results.