Recent advances in immune therapy of lymphoid cancer such as acute and chronic lymphoblastic leukaemias (ALL and CLL), lymphomas and plasma cell myeloma (PCM) have raised hopes of similar advances acute myeloid leukaemia (AML) therapy. What many fail to recognize are fundamental differences between lymphoid and myeloid neoplasms including targets of immune therapy and genetic complexity. The 1st question is: Is there immune surveillance against AML. The answer appears no. AML is not increased in persons with congenital or acquired immune deficiency nor in immune suppressed solid organ transplant recipient. Although some refer to a graft-versus-leukaemia effect seen after allotransplants there are no convincing data this is distinct from graft-versus-host disease nor that this effect operates in a non-allogeneic setting. Immune therapies of lymphoid cancers are directed against B-cell lineage targets such as CD19 or CD20. These targets are common to all B-cells and are not cancer-specific. Moreover, it is OK to destroy all B-cells, neoplastic and not, because we can replace their function with intravenous Ig. This permissiveness does not apply to targets on myeloid cells: destroying all myeloid cells, normal and neoplastic, will result in immediate death. Moreover, the genetic heterogeneity of AML (median 13 mutations) makes it difficult to develop a widely usable targeting strategy for mono- or bispecific antibodies or chimeric antigen receptor T-cells. The complexity in developing gemtuzumab ozogamicin is an example of challenges in immune therapy of AML. Fortunately, these complexities have not stopped people from trying. Recent examples include using NK-cells with some encouraging preliminary reports. Another approach is using synthetic biology. For example, inactivating CD33 on normal haematopoietic cells using CRISPR/Cas9 technology and then using an anti-CD33 antibody to eradicate CD33-positive AML cells, Diverse approaches to immune therapy of AML are displayed in the figure and will be discussed.