The method of Part I is applied to the problem of finding the characteristic function for the probability distribution of \int_0^t \sum_{jk} x_j (\tau) K_{jl}(\tau)x_l(\tau) d\tau , where x_j(\tau) denotes the j th component of a stationary n-dimensional Markoffian Gaussian process. The problem is reduced to the problem of solving 2n first-order linear differential equations with initial conditions only. For the case of constant K , the explicit solution is given in terms of the eigenvalues and the first 2n - 1 powers of a constant 2n \times 2n matrix. For the case of a symmetric correlation matrix which commutes with K , the problem is reduced to the one-dimensional case treated in Part II. For the case K_{ij}(t) = \delta_{il} \delta_{jl} e^{-t} , where the functional represents the output of a receiver consisting of a lumped circuit amplifier, a quadratic detector, and a single-stage amplifier, the solution has been obtained in a form which is more explicit than that provided by the earlier methods.