In this paper, a Nonuniform fast Fourier transform (NUFFT)-based range migration algorithm (RMA) is proposed for three-dimensional (3-D) imaging with irregular planar arrays. The proposed imaging algorithm takes advantage of NUFFT to reconstruct the signal spectra in the wavenumber domain, which overcomes the constraint of irregular spatial sampling on the utilization of Fourier-based imaging algorithms. Compared to the conventional coherent summation algorithms, i.e., backprojection and Kirchhoff migration, it is a very efficient method for 3-D image formation with nonuniform spatial sampling. To demonstrate its imaging performance, some numerical simulations are presented and discussed.Index Terms-Nonuniform fast Fourier transform (NUFFT), planar array, nonuniform spatial sampling, range migration, three-dimensional(3-D) imaging.