Shape memory polymers (SMP) demonstrate a unique ability to recover to their original shape upon application of the external stimulus after being deformed and fixed into a temporary shape. This mechanism is called shape memory effect (SME). The SMP part can be produced by injection moulding but very limited work has been done to understand the effects of moulding conditions on the SMPs. The aim of this research was to investigate the influences of selected moulding parameters on the SME. Thermoplastic Polyurethane Elastomer (TPU) with glass transition temperature Tg = 33 ºC was chosen as the SMP and three moulding parameters - injection speed, packing pressure and mould temperature were differentiated in order to produce the test parts. The combination of two levels for each process parameter was used, resulting in 8 sets of process parameters. The samples were subjected to thermomechanical (3-point bending) test, during which they were deformed at elevated temperature of (Tg + 25) ºC, fixed to a temporary shape by lowering the ambient temperature to (Tg – 5) ºC under the load and restored to their initial shape by removing the load and reapplying the heating to (Tg + 25) ºC. A systematic approach was developed for the evaluation of the SME with the use of the computed tomography (CT). The influences of the individual parameters were evaluated by the factorial analysis. Furthermore, the tensile test was performed in order to compare the recovery characteristics of the samples manufactured with the same sets of process parameters and deformed in the same ambient conditions but subjected to different form of loading. The results showed that the recovery ratio increased for high injection velocity and mould temperature and low packing pressure. The main effect plots revealed that the injection velocity had a significant influence on the recovery ratio, whereas the injection pressure and mould temperature were of minor importance.