Whilst the significance of substrate topography as a regulator of cell function is well established, a systematic analysis of the principles underlying this is still unavailable. Here we evaluate the hypothesis that surface energy plays a decisive role in substrate-mediated modulation of cell phenotype by evaluation of cell behaviour on synthetic microstructures exhibiting pronounced differences in surface energy. These microstructures, specifically cubes and walls, were fabricated from a biocompatible base polymer, poly(methyl methacrylate), by variotherm injection molding. The dimensions of the cubes were 1 μm x 1 μm x 1 μm (height x width x length) with a periodicity of 1:1 and 1:5 and the dimensions of the walls 1 μm x 1 μm x 15 mm (height x width x length) with a periodicity of 1:1 and 1:5. Mold inserts were made by lithography and electroplating. The surface energy of the resultant microstructures was determined by static contact angle measurements. Light scanning microscopy of the morphology of NT2/D1 and MC3T3-E1 preosteoblast cells cultured on structured PMMA samples in both cases revealed a profound surface energy dependence. “Walls” appeared to promote significant cell elongation, whilst a lack of cell adhesion was observed on “cubes” with the lowest periodicity. Contact angle measurements on walls revealed enhanced surface energy anisotropy (55 mN/m max., 10 mN/m min.) causing a lengthwise spreading of the test liquid droplet, similar to cell elongation. Surface energy measurements for cubes revealed increased isotropic hydrophobicity (87° max., H2O). A critical water contact angle of ≤ 80° appears to be necessary for adequate cell adhesion. A “switch” for cell adhesion and subsequently cell growth could therefore be applied by, for example, adjusting the periodicity of hydrophobic structures. In summary cell elongation on walls and a critical surface energy level for cell adhesion could be produced for NT2/D1 and MC3T3-E1 cells by symmetrical and asymmetrical energy barrier levels. We, furthermore, propose a water-drop model providing a common physicochemical cause regarding similar cell/droplet geometries and cell adhesion on the investigated microstructures.