Stereology offers a number of tools for the analysis of sections in microscopy (which usually provide only two-dimensional information) for the purpose of estimating geometric quantities, such as volume, surface area, length or number of particles (cells or other structures). The use of these tools enables recovery of the three-dimensional information that is inherent in biological tissues. This review uses the liver as a paradigm for summarizing the most commonly used state-of-the-art methods for quantitation in design-based stereology. Because it is often relevant to distinguish hyperplasia and hypertrophy in liver responses, we also focus on potential pitfalls in the sampling and processing of liver specimens for stereological purposes, and assess the existing methods for volume and number estimation. With respect to volume, we considered whole liver volume (V), volume density (VV) and so-called local volumes, including the number-weighted volume () and the volume-weighted volume (). For number, we considered the total number (N) and the numerical density (NV). If correctly applied, current stereological methods guarantee that no bias is introduced in the estimates, which will be therefore accurate; additionally, methods can be tuned for obtaining precise quantitative estimates that can reveal subtle changes in the volume or number of selected hepatic cells. These methods have already detailed the effects of some substances and specific diets on the liver, and should be routinely included in the toolbox of liver research.