Normally the starting point is the selection of a piece of tool stone that has been detached by natural geological processes, and is an appropriate size and shape. In some cases solid rock or larger boulders may be quarried and broken into suitable smaller pieces, and in others the starting point may be a piece of the debitage, a flake removed from a previous operation to make a larger tool. The selected piece is called the lithic core (also known as the "objective piece"). A basic distinction is that between flaked or chipped stone, the main subject here, and ground stone objects made by grinding. Flaked stone reduction involves the use of a hard hammer percussor, such as a hammerstone, a soft hammer fabricator (made of wood, bone or antler), or a wood or antler punch to detach lithic flakes from the lithic core. As flakes are detached in sequence, the original mass of stone is reduced; hence the term for this process. Lithic reduction may be performed in order to obtain sharp flakes, of which a variety of tools can be made, or to rough out a blank for later refinement into a projectile point, knife, or other object. Flakes of regular size that are at least twice as long as they are broad are called blades. Lithic tools produced this way may be bifacial (exhibiting flaking on both sides) or unifacial (exhibiting flaking on one side only).
Cryptocrystalline or amorphous stone such as chert, flint, obsidian, and chalcedony, as well as other fine-grained stone material, such as rhyolite, felsite, and quartzite, were used as a source material for producing stone tools. As these materials lack natural planes of separation, conchoidal fractures occur when they are struck with sufficient force; for these stones this process is called knapping. The propagation of force through the material takes the form of a Hertzian cone that originates from the point of impact and results in the separation of material from the objective piece, usually in the form of a partial cone, commonly known as a lithic flake. This process is predictable, and allows the flintknapper to control and direct the application of force so as to shape the material being worked. Controlled experiments may be performed using glass cores and consistent applied force in order to determine how varying factors affect core reduction.
It has been shown that stages in the lithic reduction sequence may be misleading and that a better way to assess the data is by looking at it as a continuum. The assumptions that archaeologists sometimes make regarding the reduction sequence based on the placement of a flake into a stage can be unfounded. For example, a significant amount of cortex can be present on a flake taken off near the very end of the reduction sequence. Removed flakes exhibit features characteristic of conchoidal fracturing, including striking platforms, bulbs of force, and occasionally eraillures (small secondary flakes detached from the flake's bulb of force). Flakes are often quite sharp, with distal edges only a few molecules thick when they have a feather termination. These flakes can be used directly as tools or modified into other utilitarian implements, such as spokeshaves and scrapers.
By understanding the complex processes of lithic reduction, archaeologists recognize that the pattern and amount of reduction contribute tremendous effect to lithic assemblage compositions. One of the measurements is the geometric index of reduction. There are two elements in this index: 't' and 'T'. The 'T' is the 'height' of maximum blank thickness and the 't' is the height of retouched scar from the ventral surface. The ratio between t and T is the geometric index of reduction. In theory this ratio shall range between 0 and 1. The bigger the number is the larger amount of lost weight from lithic flake. By using a logarithmic scale, a linear relationship between the geometric index and the percentage of original flake weight lost through retouch is confirmed. In choosing a reduction index, it is important to understand the strengths and weaknesses of each method, and how they fit to the intended research question, as different indices provide different levels of information. For example, Kuhn's geometric index of unifacial reduction (GIUR), which describes the ratio of scar height relative to the flake thickness, is highly influenced by the morphology of the flake blank which limits the applicability of this reduction index.
Alongside the various percussion and manipulation techniques described below, there is evidence that heat was at least sometimes used. Experimental archaeology has demonstrated that heated stones are sometimes much easier to flake, with larger flakes being produced in flint, for example. In some cases the heating changes the colour of the stone.