Adhesion forces between a colloid probe and some samples were consecutively measured at a single location with an atomic force microscope (AFM) in a dry environment. The outcomes show that the adhesion force depends considerably on material, contact history, and number and distribution of asperities within the contact zone. Generally, there are four different stages for the adhesion force with increasing measurement number: random behavior for the first several or tens of contacts, then increasing monotonically, later remaining stable, and finally decreasing. Usually, the measured data points are grouped into several levels. Moreover, the adhesion force jumps frequently between different levels and is more inclined to jump to a neighboring level. The level behavior was attributed to the fact that the real contact region is not exactly the same between successive measurements. The differences in the adhesion force between two neighboring levels are almost the same for one location. The adhesion force in some levels usually increases or decreases discontinuously after jumping to other levels and then jumping back. The magnitude and fluctuations of the adhesion force and the number of levels depend on the number and distribution of asperities in the contact zone.