The problem of generating optimum magnetic mirror fields at the ends of a 50-cm long theta-pinch experiment to study particle flow and loss effects has been investigated. A combination of active and passive mirrors was developed to produce 2-3:1 fields for a 23-kG pinch. Biot-Savart and potential field prediction solutions for the magnetic field distribution were compared with experimental measurements in 2-5-cm long coils for the 50-cm long pinch.