A constant matrix loadflow method is presented along with its fast version leading to the proposed fast super decoupled loadflow (FSDL) method. Gain matrices of the FSDL method are defined independent of rotation angles, and they differ only in their diagonal terms. Rotation angles are limited to the maximum value of −36°. The slack-start procedure for efficiency is also suggested. These along with a simple modification of real power mismatches at PV-nodes yield a robust and fast loadflow method. The method is tested and compared with the recent general-purpose version of the fast decoupled loadflow (GFDL) method under a wide range of simulated system operating conditions and network parameters. The proposed method has good convergence and can be more than twice as fast as the GFDL method for systems of largeR/Xratio branches.