Theories of cloaking discusses various theories based on science and research, for producing an electromagnetic cloaking device. Theories presented employ transformation optics, event cloaking, dipolar scattering cancellation, tunneling light transmittance, sensors and active sources, and acoustic cloaking.

A cloaking device is one where the purpose of the transformation is to hide something, so that a defined region of space is invisibly isolated from passing electromagnetic fields (see Metamaterial cloaking) or sound waves. Objects in the defined location are still present, but incident waves are guided around them without being affected by the object itself. Along with this basic "cloaking device", other related concepts have been proposed in peer reviewed, scientific articles, and are discussed here. Naturally, some of the theories discussed here also employ metamaterials, either electromagnetic or acoustic, although often in a different manner than the original demonstration and its successor, the broad-band cloak.

The first electromagnetic cloaking device was produced in 2006, using gradient-index metamaterials. This has led to the burgeoning field of transformation optics (and now transformation acoustics), where the propagation of waves is precisely manipulated by controlling the behaviour of the material through which the light (sound) is travelling.

Waves and the host material in which they propagate have a symbiotic relationship: both act on each other. A simple spatial cloak relies on fine tuning the properties of the propagation medium in order to direct the flow smoothly around an object, like water flowing past a rock in a stream, but without reflection, or without creating turbulence. Another analogy is that of a flow of cars passing a symmetrical traffic island – the cars are temporarily diverted, but can later reassemble themselves into a smooth flow that holds no information about whether the traffic island was small or large, or whether flowers or a large advertising billboard might have been planted on it.

Although both analogies given above have an implied direction (that of the water flow, or of the road orientation), cloaks are often designed so as to be isotropic, i.e. to work equally well for all orientations. However, they do not need to be so general, and might only work in two dimensions, as in the original electromagnetic demonstration, or only from one side, as for the so-called carpet cloak.

Spatial cloaks have other characteristics: whatever they contain can (in principle) be kept invisible forever, since an object inside the cloak may simply remain there. Signals emitted by the objects inside the cloak that are not absorbed can likewise be trapped forever by its internal structure. If a spatial cloak could be turned off and on again at will, the objects inside would then appear and disappear accordingly.

The event cloak is a means of manipulating electromagnetic radiation in space and time in such a way that a certain collection of happenings, or events, is concealed from distant observers. Conceptually, a safecracker can enter a scene, steal the cash and exit, whilst a surveillance camera records the safe door locked and undisturbed all the time. The concept utilizes the science of metamaterials in which light can be made to behave in ways that are not found in naturally occurring materials.

The event cloak works by designing a medium in which different parts of the light illuminating a certain region can be either slowed or accelerated. A leading portion of the light is accelerated so that it arrives before the events occur, whilst a trailing part is slowed and arrives too late. After their occurrence, the light is reformed by slowing the leading part and accelerating the trailing part. The distant observer only sees a continuous illumination, whilst the events that occurred during the dark period of the cloak's operation remain undetected. The concept can be related to traffic flowing along a highway: at a certain point some cars are accelerated up, whilst the ones behind are slowed. The result is a temporary gap in the traffic allowing a pedestrian to cross. After this, the process can be reversed so that the traffic resumes its continuous flow without a gap. Regarding the cars as light particles (photons), the act of the pedestrian crossing the road is never suspected by the observer down the highway, who sees an uninterrupted and unperturbed flow of cars.