The article discusses the issue of overfitting in cryo-EM structure determination, which can lead to overestimation of resolution and artifacts in the resulting maps. The authors highlight that commonly used methods, such as low-pass filtering based on suboptimal Fourier Shell Correlation (FSC) procedures, can introduce bias and inflate resolution estimates. They propose two solutions to prevent overfitting: refining two models independently to calculate gold-standard FSC curves and limiting the data used for orientation determination to a user-specified frequency. These methods are tested on simulated and real cryo-EM data sets, showing that they do not significantly affect orientation accuracy and can provide more accurate resolution estimates. The authors conclude that using gold-standard FSCs can lead to better and less noisy maps, and their procedures are straightforward to implement in existing programs.The article discusses the issue of overfitting in cryo-EM structure determination, which can lead to overestimation of resolution and artifacts in the resulting maps. The authors highlight that commonly used methods, such as low-pass filtering based on suboptimal Fourier Shell Correlation (FSC) procedures, can introduce bias and inflate resolution estimates. They propose two solutions to prevent overfitting: refining two models independently to calculate gold-standard FSC curves and limiting the data used for orientation determination to a user-specified frequency. These methods are tested on simulated and real cryo-EM data sets, showing that they do not significantly affect orientation accuracy and can provide more accurate resolution estimates. The authors conclude that using gold-standard FSCs can lead to better and less noisy maps, and their procedures are straightforward to implement in existing programs.