AbstractPerovskite solar cells have demonstrated efficiencies over 20%, but this has not been reproduced at large areas. We explore the theoretical limit to single large area perovskite solar cell efficiency, with different front conductive layers: first, the standard n‐i‐p structure with a transparent conductive electrode (TCE) at the substrate, and then structures that include a front metal grid. We model and optimize the impact of the thickness of the TCE and the dimensions of the grid elements to balance series resistance, parasitic absorption in the TCE layer, and shading losses to maximize efficiency. The results of the optimization suggest that the efficiency of the standard design can be significantly improved simply by selecting the optimum TCE thicknesses for the cell size. However, the poor scaling of the standard design prevents the fabrication of efficient large area cells. Adding a metal grid allows the efficiency of large area cells (156 mm × 156 mm) to approach that of small cells and exceed 20%. The maximum efficiency is limited by the minimum width of the metal grid elements and the sheet resistance of the metal grid material. In this context, the performances of fluorine‐doped tin oxide and indium tin oxide as TCE materials were compared, with indium tin oxide found to be superior. We also explore the limits of performance when connecting multiple cells in series to form a large area minimodule and determine the conditions when a front grid is beneficial. Design rules are presented that allow researchers to calculate the optimum cell parameters for high efficiency.