Vacuum Metal Deposition (VMD) can develop latent fingermarks through evaporation of a nano-layer of gold, followed by zinc deposition. However, effectiveness is variable, which anecdotally is dependent on donor, time and substrate. One of the possible causes could be the fingermark topography which has been very little studied. As VMD is a nano-scale process this work explores changes in fingermark topography at micro- and nanoscale on model and forensically relevant substrates. Atomic Force Microscopy (AFM) showed migration of deposited material away from latent fingermark ridges in the form of an intermediate area. This area is a uniform layer a few nm thick and is apparent for all donors, although in a range of widths from 0.6-36 microns. The intermediate areas do not always surround completely the whole of the ridges. Time of Flight Secondary Ion Mass Spectrometry (TOF-SIMS) suggested middle sized fatty acid composition of the intermediate area but this is not conclusive due to the high vacuum regime of operation of this equipment which probably affects such thin layers of material which have been found to have certain volatility. The speed of migration of material appears to vary with substrate. The extent of material migration with time on three substrates used indicated that the fastest migration is on Polyvinyl chloride (PVC) followed by Formica and polished silicon. VMD development of latent fingermarks on the same batches of the three substrates indicated that increased migration appears to increase the risk of "empty print" VMD outcome (no ridge pattern). Scanning Electron Microscopy (SEM) of vacuum-evaporated gold distribution showed different gold agglomerate behaviour on fingermark ridges, intermediate area and model polished silicon substrate between ridges. This suggests that the intermediate area is the cause for empty print VMD development and will have substantial effects on other development methods.