The use of human pluripotent stem cells (hPSCs) in regenerative medicine has great potential. However, it is important to exclude that these cells can undergo malignant transformation, which could lead to the development of malignant tumours. This property of hPSCs is currently being tested using the teratoma assay, through which cells are injected into immunodeficient mice. Transplantation of stem cells in immunocompromised recipient animals certainly has a much higher incidence of tumour formation. On the other hand, the results obtained in immunodeficient mice could indicate a risk of tumour formation that is practically not present in the human immunocompetent recipient. The presence of a humanised immune system might be more representative of the human situation; therefore, we investigated if the demonstrated malignant features of chosen and well-characterised stem cell lines could be retrieved and if new features could arise in a humanised mouse model. Hu-CD34NSGTM (HIS) mice were compared side by side with immunocompromised mice (NSG) after injection of a set of benign (LU07) and malignant (LU07+dox and 2102Ep) cell lines. Analysis of the tumour development, histological composition, pathology evaluation, and malignancy-associated miRNA expression levels, both in tumour and plasma samples, revealed no differences among mouse groups. This indicates that the HIS mouse model is comparable to, but not more sensitive than, the NSG immunodeficient model for studying the malignancy of stem cells. Since in vivo teratoma assay is cumbersome, in vitro methods for the detection of malignancy are urgently needed.