In physical chemistry, supersaturation occurs with a solution when the concentration of a solute exceeds the concentration specified by the value of solubility at equilibrium. Most commonly the term is applied to a solution of a solid in a liquid, but it can also be applied to liquids and gases dissolved in a liquid. A supersaturated solution is in a metastable state; it may return to equilibrium by separation of the excess of solute from the solution, by dilution of the solution by adding solvent, or by increasing the solubility of the solute in the solvent.

Early studies of the phenomenon were conducted with sodium sulfate, also known as Glauber's Salt because, unusually, the solubility of this salt in water decreases with increasing temperature past 33°C. Early studies have been summarised by Tomlinson. It was shown that the crystallization of a supersaturated solution does not simply come from its agitation, (the previous belief) but from solid matter entering and acting as a "starting" site for crystals to form, now called "seeds" (for more information, see nucleation). Expanding upon this, Gay-Lussac brought attention to the kinematics of salt ions and the characteristics of the container having an impact on the supersaturation state. He was also able to expand upon the number of salts with which a supersaturated solution can be obtained. Later Henri Löwel came to the conclusion that both nuclei of the solution and the walls of the container have a catalyzing effect on the solution that cause crystallization. Explaining and providing a model for this phenomenon has been a task taken on by more recent research. Désiré Gernez contributed to this research by discovering that nuclei must be of the same salt that is being crystallized in order to promote crystallization.

Furthermore, in 1950, Victor K. LaMer proposed another theory for nucleation, in which he described the nucleation and growth of sulfur nuclei in a solution where a chemical reaction provided a constant inflow of molecularly dissolved sulfur. This theory, however, is not confined to this specific case and can be generalised as shown in LaMer’s diagram, provided in the second figure of this section.

In section (I), the concentration of solute grows linearly, as it is formed (or added) to the solution. Upon reaching ${\displaystyle c_{L}^{eq}\,\!}$, it will become saturated, but it won’t start depositing solute right away. Instead, it will keep absorbing it, becoming supersaturated.

In section (II), concentration reaches critical saturation levels, ${\displaystyle c_{min}\,\!}$, when solute crystals begin nucleating. The appearance of nuclei partially relieves the supersaturation, at least rapidly enough that the rate of nucleation falls almost immediately to zero. The system rapidly reaches a balance between the solute supply and the consumption rate for the nucleation and its growth, slowing down the increase in its concentration. After reaching the peak, the curve declines owing to the increasing consumption of the solute for the growth of nuclei and reaches again the critical level of nucleation, ${\displaystyle c_{min}\,\!}$, ending the nucleation stage. Given optimal conditions, having the solute be introduced to the solution very steadily while keeping the system free from perturbations and nucleation seeds, the maximum concentration that can be achieved in this way is defined as ${\displaystyle c_{max}\,\!}$.

In section (III), the supersaturation becomes too low for any more crystals to nucleate, so no new crystals are formed. However, as the solution is still supersaturated, the existing crystals grow by solute diffusion. As time passes by, the growth rate of the crystal equals the rate of solute supply, so the concentration converges to the saturation value ${\displaystyle c_{L}^{eq}\,\!}$.

A solution of a chemical compound in a liquid will become supersaturated when the temperature of the saturated solution is changed. In most cases solubility decreases with decreasing temperature; in such cases the excess of solute will rapidly separate from the solution as crystals or an amorphous powder. In a few cases the opposite effect occurs. The example of sodium sulfate in water is well-known and this was why it was used in early studies of solubility.

Recrystallization is a process used to purify chemical compounds. A mixture of the impure compound and solvent is heated until the compound has dissolved. If there is some solid impurity remaining it is removed by filtration. When the temperature of the solution is subsequently lowered it briefly becomes supersaturated and then the compound crystallizes out until chemical equilibrium at the lower temperature is achieved. Impurities remain in the supernatant liquid. In some cases crystals do not form quickly and the solution remains supersaturated after cooling. This is because there is a thermodynamic barrier to the formation of a crystal in a liquid medium. Commonly this is overcome by adding a tiny crystal of the solute compound to the supersaturated solution, a process known as "seeding". Another process in common use is to rub a rod on the side of a glass vessel containing the solution to release microscopic glass particles which can act as nucleation centres. In industry, centrifugation is used to separate the crystals from the supernatant liquid.