Single molecule localization based super resolution microscopy relies on precise and accurate localization of a large number of single molecules. However, the necessity of accumulating large numbers of localization estimates limits the time resolution to typically seconds to minutes.Two of the major limitations are the acquisition speed and the photon budget. Replacing EMCCD cameras which are usually implemented in such experiments with recently introduced SCMOS cameras results in a leap in both acquisition speed and effective quantum efficiency. However, the intrinsic pixel-dependent Gaussian noise of the SCMOS cameras introduces localization artifacts and greatly reduces the reliability of the results.Here, we present a set of specially designed statistics-based algorithms that allows for the first time to fully characterize an SCMOS camera and perform unbiased and precise localization analysis. Using this method we demonstrate Cramer-Rao lower bound-limited single molecule localization with an SCMOS camera. Combining this with a recently developed multi-emitter fitting algorithm and optimized imaging condition, we show that this technique shortens the typical acquisition time for fixed samples by up to two orders of magnitude without compromising the field of view. Furthermore, we demonstrate video-rate super-resolution microscopy of live cells by monitoring the transient clustering events of transferrin receptors with a reconstructed frame rate of 32 fps.The here presented method allows to replace EMCCD cameras with SCMOS technology and record faster and more precise super-resolution images without compromises.