A computational method is described that allows the measurement of the signal-to-noise ratio and resolution of a three-dimensional structure obtained by single particle electron microscopy and reconstruction. The method does not rely on the availability of the original image data or the calculation of several structures from different parts of the data that are needed for the commonly used Fourier Shell Correlation criterion. Instead, the correlation between neighboring Fourier pixels is calculated and used to distinguish signal from noise. The new method has been conveniently implemented in a computer program called RMEASURE and is available to the microscopy community.