The integrated planar solid oxide fuel cell designed by Rolls-Royce pic consists of a fuel electrode and an air electrode separated by an ion-conducting solid electrolyte. The electrolyte must exhibit ionic conductivity and be non-permeable to the cell gases. The other cell components set limits for the firing temperature of the electrolyte, thus restricting its densification. Further, the use of a thin constrained layer, fabricated via screen-printing, makes the separation of the fuel and oxidant much more challenging than in unconstrained planar fuel cell designs. Thus, the aim of this study was to investigate the influence of the screen-printing and sintering processes on the microstructure of the electrolyte and ultimately on the permeability or gas flow rate. The ceramic material used throughout this study was zirconia doped with 3-mol% yttria formulated into a screen printable ink and deposited onto a pre-sintered substrate. The ink was deposited through mesh sizes ranging in thickness from 50 to 215 mum. The printed layer was then dried and additional layers added up to a maximum of four applications. The resulting laminate was sintered using a conventional sintering scheme. The thickness, relative density and grain size of the sintered electrolyte were recorded. Increasing sintered layer thickness showed an associated reduction in relative density and grain size. The conventional sintering profile was modified by changing the heating rate, sintering temperature and sintering time. From this part of the study a baseline fabrication process was established. This baseline used a 165 mesh size with three print-dry applications, a heating rate of 10°C min-1, a sintering temperature of 1450°C and a sintering time of 10 hours. The conventional sintering profile was modified to incorporate coarsening, two-step sintering and a combination of these denoted as 'three-step' sintering. Layer thickness, relative density, grain size distribution and pore size distribution were determined. It was possible to achieve full density with all three types of sintering profile. Grain growth could not be avoided in the constrained thick-films but the pore size distribution was narrower for the modified sintering profiles compared with that for conventional sintering. Gas leakage testing showed that the effect of microstructure on the gas flow through the electrolyte was not a function of a single variable. Rather, it was a combination of at least four parameters (the ratio of grain size to sintered layer thickness, relative density, pore size range and number of pores per unit area). Only if all four parameters were within certain defined boundary conditions was the gas flow minimised. Although the lowest gas flow rate measured was still just outside of the design requirement for an industrial application, the associated sintering schedule used lower temperatures than currently used to densify the electrolyte and there is scope for further optimisation.