Zinc alloy castings are usually assembled together or mounted by screwed steel fasteners, and are tightened to a predetermined torque to develop the required tensile preload in the fastener. Due to relaxation processes in the castings, creep may cause a partial preload loss at an elevated temperature. The equipment used for load relaxation tests consists of a load-monitoring device, an oil bath, and a data-acquisition system. A load cell monitoring device is used to monitor the load loss in an ISO-metric M6 x 1 steel screw set into sand castings made from alloys No. 3, No. 5 and No. 2 and tightened to produce an initial preload of 6 kN. The castings were held at constant temperature in the range 80-120 degrees C in an oil bath. The oil bath maintains the desired test temperature throughout the experiment. All tests were conducted for periods of upto 160h. For all alloys, the initial load loss was high, decreasing gradually with time, but not ceasing. The load loss increased rapidly with test temperature, and almost all of the relaxation curves approximated to a logrithmic decay of load with time. Alloy No. 2 had the best resistance to load loss, with No. 5 next and No. 3 worst at all temperatures. The lower resistance to relaxation of alloy No. 3 was mainly due to the lower relaxation strength of copper-free primary dendrites, whereas in alloys No. 5 and No. 2, the higher copper contents contribute greatly to their relaxation strength in the form of second-phase particles.