Areal spectral analysis is applied to achieve proper frequency bandwidth for topography measurements from machined surfaces. The effect of aliasing on the spectrum of surface features due to a large sample spacing is evaluated. The characteristics of low-frequency components of a surface are also investigated. The low-frequency limit of measurement for obtaining the genuine features of a surface can be decided from the spectral analysis of low-frequency components of the surface. The guidelines for topography measurement are proposed for selecting an appropriate sampling bandwidth of surfaces with due consideration of limited computer memory. The guidelines developed are applied to study the general patterns of spectra from a range of isotropic engineering surfaces. An example is given to demonstrate the selection of a temporary sample spacing for a further measurement searching for the low-frequency cut-off of a surface, which may be in turn used to determine the high-frequency limit with an estimate of aliasing influence. In spectral analysis, circular spectra and/or accumulation spectra have been shown to be suitable for the analysis of sample surfaces from isotropic machining processes.