In this study we measured the adhesion forces between atomic force microscope (AFM) tips or particles attached to AFM cantilevers and different solid samples. Smooth and homogeneous surfaces such as mica. silicon wafers. or highly oriented pyrolytic graphite, and more rough and heterogeneous surfaces such as iron particles or patterns of TiO2 nanoparticles on silicon were used. In the first part, we addressed the well-known issue that AFM adhesion experiments show wide distributions of adhesion forces rather than a single value. Our experiments show that variations in adhesion forces comprise fast (i.e., from one force curve to the next) random fluctuations and slower fluctuations, which occur over tens or hundreds of consecutive measurements. Slow fluctuations are not likely to be the result of variations in external factors such as lateral position, temperature, humidity, and so forth because those were kept constant. Even if two solid bodies are brought into contact under precisely the same conditions (same place, load, direction, etc.) the result of such a measurement will often not be the same as that of the previous contact. The measurement itself will induce structural changes in the contact region, which can change the value for the next adhesion force measurement. In the second part, we studied the influence of humidity on the adhesion of nanocontacts. Humidity was adjusted relatively fast to minimize tip wear during one experiment. For hydrophobic surfaces, no signification change in adhesion force with humidity was observed. Adhesion force versus humidity curves recorded with hydrophilic surfaces either showed a maximum or continuously increased. We demonstrate that the results can be interpreted with simple continuum theory of the meniscus force. The meniscus force is calculated based on a model that includes surface roughness and takes into account different AFM tip (or particle) shapes by a two-sphere model. Experimental and theoretical results show that the precise contact geometry has a critical influence on the humidity dependence of the adhesion force. Changes in tip geometry on the sub-10-nm length scale can completely change adhesion force versus humidity curves. Our model can also explain the differences between earlier AFM studies, where different dependencies of the adhesion force on humidity were observed.