Firstly, I would like to say, this contribution is written only for this Dedicated Issue of CCCC, only for my sweet boss, dear friend and my heavy criticZrzek. Time to time I attended some conferences focused on modeling, simulations and I understood this point badly. Many renowned professors spoke about some playing with balls and only gradually I have found out that "playing with small balls" could be useful, important and had many applications. To delight my dear Zrzek, I started to look intensively for some suitable "ball problem" for me. Thanks to my dear colleagues, co-authors of this paper, I got a chance for "balling". I have chosen the special problem of gold nano-clusters and nano-layers sputtered on solid surfaces.The gold density enters some indirect determination of gold properties as an input parameter. Reported gold density of 19 320 kg m -3 at standard conditions is correct for gold bulk or macro-sized specimens. For nano-sized specimens (nano-clusters, nano-layers), however, this standard density value need not be applicable. This work deals with gold density as a function of the nano-structure size. The calculated density was compared with experimental one performed for gold sputtered onto glass or poly(ethylene terephthalate) substrates. Calculated gold density was found to increase from 11 249 to 19 304 kg m -3 when the characteristic dimension of gold nano-structures increases from 0.7 to 82 nm. At the same time the free volume fraction in nano-structures decreases from 57 to 26%.Vacuum-deposited metal clusters and thin films play an important role in various fields of technology 1 . Gold in the form of thin films is nowadays used in a vast range of applications such as micro-and nano-electromechanical systems (MEMS and NEMS) 2 , sensors 3 , electronic textiles 4 , bioengineering 5 , generators of non-linear optical properties 6 or devices for surface-enhanced Raman scattering 7 .