The testing procedures and reported tribological performance of polymer gears, which are seeing ever-increasing use by industry, still lack consistency and detail. One of the key parameters that affect a polymer's mechanical properties, interface contact conditions and, consequently, the wear and fatigue behaviour, is the temperature. Temperature is well known to have a critical influence on all polymers, much more so than on metals or other materials. However, the temperatureseven for the same load and velocityvary greatly with testing devices, gear sizes and shapes, surrounding environment, as do the cooling and heating rates. Moreover, the temperature, either root, flank or bulk, also varies with the operating conditions, meaning that even in the same S-N curve data at different stresses we actually obtain different temperatures. This suggests that the results from different tests and even different load levels, which are not temperature controlled, cannot be directly compared due to the important influence of the temperature on the polymers. This work presents the results of POM gears tested against steel gears under well-controlled temperature conditions, providing various S-N curves obtained at 30 ˚C, 50 ˚C and 70 ˚C that were kept constant at the gear root for three torque values (1.0 Nm, 1.2 Nm and 1.4 Nm), and these are compared to data obtained under conditions that are typically considered to be "room temperature", which is in fact an uncontrolled temperature. The results confirm the important effect of the temperature on the fatigue life of POM gears, i.e., when kept at low temperature (30 °C), the fatigue life of POM gears is greatly improved.