Nanoindentation has become a valuable tool in wood science. It enables to examine the mechanical properties of the wood cell walls, which are polymeric, multi-layered structures with typical thicknesses of a few micrometers. Despite the intensive use of the method for the characterization of wood cell walls, it is not entirely clear yet how the measurement results may be affected by the way the sample is prepared and the indentation is carried out. This manuscript contributes to clarify these issues, by presenting indentation data for a variety of sample preparation techniques and indentation protocols, and by critically evaluating the observed differences of the obtained indentation moduli and hardnesses. Investigations covered the effect of different embedding materials, including testing of non-embedded cell walls, and of repeated exposure to high temperatures during harsh drying before the indentation test. Moreover, potential edge effects were studied when the indentation size approaches the width of the individual cell wall layers. Using different embedding materials as well as testing non-embedded cell walls did not lead to significant changes in the measured properties. Due to damage during the sample preparation, non-embedded cell walls tend to show substantially higher experimental scatter. Repeated drying prior to embedding had no significant effect on the resulting moduli and hardnesses. Finally, it was found that reasonable mechanical properties can be extracted from the cell corner middle lamella (CCML), even when the size of the indent approaches the diameter of the CCML.