Practical thin‐film dc superconducting quantum interference devices (SQUIDs) are often built with an integrated input coil which leads to the formation of various parasitic elements. We have designed and fabricated several square washer‐shaped dc SQUID test devices with various amounts of parasitic capacitance formed by covering the washer slits with superconducting coverplates of various widths. In the simplest approximation, the parasitic capacitance Cp appears entirely across the Josephson junctions; in a better approximation, Cp divides the total inductance into two parts, an effective SQUID loop inductance L and a small parasitic inductance Lp. For both models, noise‐free and noise‐rounded current‐voltage and voltage‐flux characteristics are calculated for various values of the ratios L/Lp and Cp/C, where C is the Josephson junction capacitance. The SQUID dynamics are significantly affected by Cp and Lp. The energy resolution ϵ at first deteriorates with increasing Cp for Cp/C ≲ 2, but as Cp/C becomes large, ϵ improves to less than twice the value for Cp/C ≊ 0. The optimal points of operation for large Cp, however, are found at voltages above the low‐frequency resonance determined by L and Cp where the flux‐to‐voltage transfer function is low. Furthermore, for devices with Cp/C ≫ 1 and Lp≠0, the double‐loop geometry leads to an additional resonance determined by Lp and C that appears at high frequencies. The energy resolution inside this high‐frequency resonance is nearly independent of Cp and comparable to the energy resolution of a device having the same total inductance but with Cp/C ≪ 1. Extensive noise mappings on several - different test devices have been carried out using a computer‐controlled measurement system. The measured characteristics and noise performance are shown to be in good agreement with the simulations.