Scanning ion conductance microscopy (SICM) can be used to obtain the surface topography of the sample under noncontact conditions. It can realize the nondestructive imaging of soft samples, e.g., living cells in physiological liquid environment. However, there is smearing phenomenon in the SICM image when using the continuous feedbackcontrol scanning mode which is found through a large number of experimental results. It causes the image distortion and limits the scan speed. To solve this problem, the SICM imaging principle is analyzed. It is concluded that the highly nonlinearity of the approach curve is the main reason for this phenomenon, and an adaptive control method based on currenterror compensation model for SICM is proposed in this study. The main idea is to establish a currenterror compensation model, predict the current scanning point position by using the scanning data of the last line as the prior knowledge, and then put it into the compensation model to calculate the new currenterror as the system controlled variable. Finally, the performance of scan images for the standard grating under the new and old control algorithms is compared. The experimental results show that the new algorithm can effectively solve the smearing phenomenon at a certain degree of scanning speed and significantly reduce the image distortion. It provides an effective technical method to improve the image quality and imaging speed of SICM system.