This paper presents experimental and analytical studies of AFM nanoindentation as a method to determine microhardness of hard thin films. Indentations are performed on Au, Si, and DLC using triangular pyramidal diamond probes to determine microhardness. We examined the effects of indentation force and three different methods to measure the indentation area (direct area measurement for a triangular indentation, size analysis of an inverted AFM image of the indentation, and prediction of area from indentation depth based on a tip shape function). The responses are indentation depth, projection area, and microhardness. Relationship of responses with indentation force is examined. At a low depth range, microhardness based on pyramidal shape function is erroneously higher than the other two methods. However, all three methods generally agree with each other when indentation force exceeds 16μN. Size analysis of inverted images gives more consistent microhardness with the least variability over the whole force (4∼30μN) or depth (0.5∼70 nm) range. Tip shape models are developed to predict microhardness. The ellipsoidal tip model is a good approximation at low depth ranges, while the pyramidal model works better for deeper indentations. The force vs area curves are also significantly nonlinear which can distort the hardness measurements.