CTFFIND4 is an improved version of the CTFFIND program for estimating objective lens defocus parameters from transmission electron micrographs. It is significantly faster and more suitable for modern imaging technologies like dose fractionation and phase plates. The algorithm models the microscope's contrast transfer function (CTF) as a function of spatial frequency, allowing for accurate defocus and astigmatism estimation. CTFFIND4 uses a simplified CTF model that reduces computational cost while maintaining accuracy. It also includes a method to assess the quality of the fit as a function of spatial frequency, which can define the highest resolution at which CTF oscillations are successfully modeled. The program allows users to input micrograph files and other parameters, with default values provided for guidance. It supports processing of dose-fractionated movies and micrographs with phase plates, which introduce variable phase shifts. CTFFIND4 includes a feature to estimate phase shifts from the data, improving the accuracy of defocus and astigmatism parameters. The algorithm involves computing the amplitude spectrum, background subtraction, and a scoring function to maximize the similarity between theoretical CTF functions and the experimental data. It uses a 3D conjugate-gradient method to refine the estimates of defocus and astigmatism parameters. CTFFIND4 also provides diagnostic outputs, including a diagnostic image and 1D profiles of the amplitude spectrum and CTF fit, which help assess the quality of the fit. Benchmarking shows that CTFFIND4 is significantly faster than CTFFIND3, with speedups of up to 10-fold in some cases. It maintains similar accuracy in estimating defocus parameters, with discrepancies usually below 1%. The program also provides a measure of the quality of the fit, using a threshold of CC_fit = 0.75 to determine the highest resolution at which Thon rings are reliably fit. CTFFIND4 is suitable for use with modern imaging technologies and provides accurate defocus and astigmatism estimates, making it a valuable tool for electron microscopy.CTFFIND4 is an improved version of the CTFFIND program for estimating objective lens defocus parameters from transmission electron micrographs. It is significantly faster and more suitable for modern imaging technologies like dose fractionation and phase plates. The algorithm models the microscope's contrast transfer function (CTF) as a function of spatial frequency, allowing for accurate defocus and astigmatism estimation. CTFFIND4 uses a simplified CTF model that reduces computational cost while maintaining accuracy. It also includes a method to assess the quality of the fit as a function of spatial frequency, which can define the highest resolution at which CTF oscillations are successfully modeled. The program allows users to input micrograph files and other parameters, with default values provided for guidance. It supports processing of dose-fractionated movies and micrographs with phase plates, which introduce variable phase shifts. CTFFIND4 includes a feature to estimate phase shifts from the data, improving the accuracy of defocus and astigmatism parameters. The algorithm involves computing the amplitude spectrum, background subtraction, and a scoring function to maximize the similarity between theoretical CTF functions and the experimental data. It uses a 3D conjugate-gradient method to refine the estimates of defocus and astigmatism parameters. CTFFIND4 also provides diagnostic outputs, including a diagnostic image and 1D profiles of the amplitude spectrum and CTF fit, which help assess the quality of the fit. Benchmarking shows that CTFFIND4 is significantly faster than CTFFIND3, with speedups of up to 10-fold in some cases. It maintains similar accuracy in estimating defocus parameters, with discrepancies usually below 1%. The program also provides a measure of the quality of the fit, using a threshold of CC_fit = 0.75 to determine the highest resolution at which Thon rings are reliably fit. CTFFIND4 is suitable for use with modern imaging technologies and provides accurate defocus and astigmatism estimates, making it a valuable tool for electron microscopy.