The effects of solvent composition and degree of reaction prior to film formation leading to the formation of a biphasic microtopographical surface in a crosslinked siloxane-urethane coating system were explored. For the solvent composition study, a D-optimal mixture design study was carried out using methyl n-amyl ketone (MAK), toluene, ethyl 3-ethoxypropionate (EEP), butyl acetate (BA) and isopropyl alcohol (IPA) as solvents. The study revealed that the presence of slow evaporating solvents MAK, EEP and the absence of fast evaporating solvent IPA in the solvent composition with a minimal amount of BA favored formation of a structured surface. Control over the domain size could be obtained by varying the MAK:EEP ratio in solvent compositions having a fixed amount of BA. The effect of mixing time on the formation of surface domains was studied. At short mixing times (< 2 h) and long mixing times (> 7 h), no surface phase separation is observed, while at intermediate times a 'window' was found where surface microdomains of similar size are generated. Doubling the level of catalyst halves the mixing time required to generate surfaces with microdomains. Scanning electron microscopy (SEM) studies with energy dispersive X-ray mapping and dynamic mechanical analysis (DMA) were done in order to understand the development of the phases in the PDMS-polyurethane system. (c) 2006 Elsevier Ltd. All rights reserved.