Abstract A novel test system for fatigue crack growth (FCG) testing with cracked round bar (CRB) specimens was developed and implemented, utilizing a commercially available linear-torsion electro-dynamic test machine, several specifically designed test fixtures and devices, and a unique test and data reduction procedure. The test arrangement includes two different cyclic testing devices, which allow for tests at adjustable test temperatures and under predefined environmental conditions with variable gaseous or liquid environmental media. The test system and procedure allow for a quasi-automatic determination of crack growth rates via a self-designed optical crack length measurement device and tests air or other transparent gaseous or liquid environmental media. To account for the frequently observed asymmetric crack growth in CRB specimens, a special technique has been established, in which the entire CRB specimen is rotated in the test machine to enable measurements of crack lengths at various circumferential positions. To demonstrate the potential of the test system, high-density polyethylene (PE-HD) was investigated at an elevated temperature of 80 °C and in different environmental media (air, non-chlorinated water and chlorinated water with 5 mg/l free chlorine). When using the crack length mean value derived from crack length measurements from various circumferential positions, the fatigue crack growth (FCG) resistance of PE-HD at 80 °C in air was found to be independent of the degree of crack growth asymmetry. As to the environmental influence, chlorinated water with a chlorine content of 5 mg/l free chlorine was found to represent a more critical environmental condition than non-chlorinated water. Fatigue crack growth curves in air exhibited a significantly slower slope than the other two environments, leading to a crossover of the FCG curves with a tendency for higher and lower crack growth rates, respectively, at low and high stress intensity factor levels.