Unlike traditional wet-adhesives, dry adhesives are bioinspired solid polymers that owe their adhesion to hierarchical structures. Gecko’s are the most recognizable living creatures relying on this concept to climb walls. A close look at the gecko’s fingers reveals a hierarchical system of micro-and nanoscale filaments. These ensure excellent adaptability and contact to (rough) surfaces and the formation of a large amount of close-range van der Waals surface interactions. The reversible nature of these bonds allows for reversible and repeatable adhesion. This hierarchical system has been identified as the main reason for dry adhesion and has been object of intense research to manufacture man-made solid adhesives. Despite the many efforts, there are some unclear aspects resulting from contradictory scientific reports and a strong focus on a particular polymer chemistry used to make the hierarchical structures. This has left the effects of the material choice on dry adhesion largely neglected. In this research project, we aimed to shed some light on the role of different polymer architecture features on dry adhesion. Most available research has focused on the use of commercial siloxane elastomers that offer almost no control on the polymer synthesis. Instead, we opted to use thermoplastic polyurethane chemistry due to the large versatility this chemistry offers in terms of modification of relevant polymer architecture features. In the absence of available works on thermoplastics a new manufacturing process of such hierarchical structures had to be developed relying on heating, pressure and vacuum to respectively melt the polymer, overcome the high viscosity and impede oxidation. We studied the effects of the polymer architecture and properties on adhesion using a single micropillar architecture and varying the chemical composition (polyol length, aromatic content, hard/soft block ratio), the testing temperature and the pull-off speed. As to the polymer architecture, the best results were found with a short polyol and the lowest hard block fraction that still guaranteed structural integrity. Next to that, it was found that having aromatic rings in the hard segments was crucial, likely due to its beneficial effect on nanophase separation. With those compositions, values exceeding those of state-of-the-art dry adhesives were found, with a maximum of 440 kPa at the Tg and high retraction speeds. Furthermore, we show that while existing models are valid for thermoplastic polyurethanes well below their Tg, once they exhibit viscoelastic behaviour, the loss factor is a much more reliable indicator of performance than the reported stiffness. The effect of the surface energy was also evident but minimal compared to the mechanical properties of the polymers.