Segmental rotor switched reluctance machines have been demonstrated to produce significantly more torque than conventional switched reluctance machines because each coil is able to link more magnetic flux. In previous work, the range of topologies studied has been relatively restricted; there has been no consideration of the range of stator tooth and rotor segment number options, and the designs have not been formally optimized in any way. This paper seeks to address this: Using finite-element toolsets for optimization, it describes the resulting machines—uniquely featuring a greater number of rotor segments than stator teeth. This paper assesses the credible tooth and segment options, with each option optimized to maximize torque capability. From this assessment, conclusions are drawn, which contrast each machine's performance based on criteria including torque production and loss. In order to validate this optimization process, a prototype machine is built, and its performance is compared with these predictions.