Test assays capable of providing quantitative characterization of the contraction of cardiac and smooth muscle cells are of great need for drug development and screening. Several methodologies have been proposed for achieving measurement of cell contractile stress or force, however almost all rely on optical methods to detect contraction. Recently, we proposed a test assay method based on the cell-induced deformation of thin-film, elastomeric, capacitive sensors. The method uses an electrical (capacitive) read-out enabling facile up-scaling to a large number of devices working in parallel for high-throughput measurements. We present here a model for the prediction and optimization of sensor performance. Our model shows the following trends: a) a cell region ratio of approximately 0.75 of the culture well radius produces the largest change in capacitance for a given cell contractile stress, b) the change in capacitance generated by cell contraction increases as the Young’s modulus, sensing layer thickness and electrode thicknesses of the sensor decrease, following an inverse relationship. A prototype device is fabricated and characterized in cell culture conditions. Mean standard deviations as lows as 0.2 pF are achieved (<0.05% of the initial sensor capacitance), representing a minimum detectable cell stress of 1.2 kPa, as predicted by our model. This sensitivity is sufficient to measure the contractile stress of smooth and cardiac muscle cell monolayers as reported in the literature.