Polynomial-phase signals have attracted significant interest due to their applicability to radar, sonar, geophysics, and radio communication. In this paper we introduce a new technique for estimating the parameters of polynomial phase signals. The parameters are estimated by performing phase unwrapping in a least squares manner. The least squares problem is formulated as a nearest lattice point problem that can be solved using existing techniques. The statistical performance of the new estimator is excellent when compared with popular existing estimators such as those based on the discrete polynomial phase transform.