The four-vector electromagnetic potential is defined in SI units (and gaussian units in parentheses) as

in which φ is the electrical potential, and A is the magnetic potential, a vector potential.

It is useful to group the potentials together in this form because is a Lorentz covariant vector, meaning that it transforms in the same way as the spacetime coordinates (ct, x) under transformations in the Lorentz group: rotations and Lorentz boosts. As a result, the length

is the same in every inertial reference frame.

Often, physicists employ the Lorenz gauge condition to simplify Maxwell's equations as:

where J is the four-current,

and

is the d'Alembertian operator.

In terms of the scalar and vector potentials, this last equation becomes:

For a given charge and current distribution, ρ(x, t) and J(x, t), the solutions to these equations in SI units are

,

where τ = t - |x-x′|/c is the retarded time . This is sometimes also expressed with ρ(x′, τ) = [ρ(x′, t)], where the square brackets are meant to indicate that the time should be evaluated at the retarded time. Of course, since the above equations are simply the solution to an inhomogeneous differential equation, any solution to the homogeneous equation can be added to these to satisfy the boundary conditions. These homogenous solutions in general represent waves propagating from sources outside the boundary.

When the integrals above are evaluated for typical cases, eg of an oscillating current (or charge), they are found to give both a magnetic field component varying as r ^{- 2} (the induction field) and a component decreasing as r ^{- 1} (the radiation field).