Modern scanning transmission electron microscopy (STEM) enables imaging and microanalysis at very high magnification. In the case of aberration-corrected STEM, atomic resolution is readily achieved. However, the electron fluxes used may be up to three orders of magnitude greater than those typically employed in conventional STEM. Since specimen contamination often increases with electron flux, specimen cleanliness is a critical factor in obtaining meaningful data when carrying out high magnification STEM. A range of different specimen cleaning methods have been applied to a variety of specimen types. The contamination rate has been measured quantitatively to assess the effectiveness of cleaning. The methods studied include: baking, cooling, plasma cleaning, beam showering and UV/ozone exposure. Of the methods tested, beam showering is rapid, experimentally convenient and very effective on a wide range of specimens. Oxidative plasma cleaning is also very effective and can be applied to specimens on carbon support films, albeit with some care. For electron beam-sensitive materials, cooling may be the method of choice. In most cases, preliminary removal of the bulk of the contamination by methods such as baking or plasma cleaning, followed by beam showering, where necessary, can result in a contamination-free specimen suitable for extended atomic scale imaging and analysis. Modern scanning transmission electron microscopy (STEM) enables imaging and microanalysis at very high magnification. In the case of aberration-corrected STEM, atomic resolution is readily achieved. However, the electron fluxes used may be up to three orders of magnitude greater than those typically employed in conventional STEM. Since specimen contamination often increases with electron flux, specimen cleanliness is a critical factor in obtaining meaningful data when carrying out high magnification STEM. A range of different specimen cleaning methods have been applied to a variety of specimen types. The contamination rate has been measured quantitatively to assess the effectiveness of cleaning. The methods studied include: baking, cooling, plasma cleaning, beam showering and UV/ozone exposure. Of the methods tested, beam showering is rapid, experimentally convenient and very effective on a wide range of specimens. Oxidative plasma cleaning is also very effective and can be applied to specimens on carbon support films, albeit with some care. For electron beam-sensitive materials, cooling may be the method of choice. In most cases, preliminary removal of the bulk of the contamination by methods such as baking or plasma cleaning, followed by beam showering, where necessary, can result in a contamination-free specimen suitable for extended atomic scale imaging and analysis.