Diffusion is the most common surface reaction and, thus, arguably the most important for both surface reactivity and adlayer formation (e.g. thiol-based self-assembled monolayers) Yet it is hard to study by surface averaging techniques such as photoelectron- or infrared-spectroscopy, as the initial and final states of diffusion are very similar and the actual time spent in transit is far too short to have any spectroscopic significance. I will present a range of scanning tunneling microscopy (STM) based approaches towards the understanding of the diffusion of small molecules on metal surfaces. The most facile approach is the acquisition of time-lapsed images that allow tracking of molecular movement on surface. This approach requires tuning of the STM temperature to a range, in which the molecular movement does not exceed the image acquisition rate of STM of a few seconds. Hence, it does not allow addressal of conditions that are relevant for chemical surface reactivity, which occurs at higher temperature. In my talk I will detail atomically-resolved measurements of laser-induced diffusion that are capable of operation at higher adsorbate excitation level and I will complement their results with findings derived from STM-based translation of individual molecules.