Thirteen sets of dished accelerator grids were tested on five different 30-cm diameter bombardment thrusters to evaluate the effects of grid geometry variations on thruster discharge chamber performance. The dished grid parameters varied were: grid-to-grid spacing, screen and accelerator grid hole diameter, screen and accelerator open area fraction, compensation for beam divergence losses, and accelerator grid thickness. Also investigated were the effects on discharge chamber performance of main magnetic field changes, magnetic baffle current, cathode pole piece length and cathode position.