We propose a computationally efficient algorithm designed to address the curse of increasing dimensions found in the Skewed Kalman Filter. The algorithm’s accuracy and efficiency are substantiated through a comprehensive simulation study encompassing both univariate and multivariate state-space models. We demonstrate applicability by estimating a multivariate dynamic Nelson-Siegel term structure model and a New Keynesian DSGE model on US data with Maximum Likelihood. In both applications, the results reveal a strong preference for a skewed error term distribution.