Investigation of the mechanism of relapse in patients receiving stem cell rescue as therapy for malignant disease has been facilitated by gene marking studies. These studies have shown the marker gene to be present in malignant cells in the patient at the time of relapse, indicating that infused stem cells can contribute to disease recurrence. As normal progenitor cells are also marked and can be tracked in vivo, these studies have also helped us learn how haemopoietic stem cells respond to manipulation, for example with growth factors. Second generation studies with multiple, modified vectors are beginning to provide information about a wider variety of clinical and biological issues, including the efficacy of purging. Although marker studies have been useful for haematological malignancy and for neuroblastoma, they are hampered by the low efficiency of marking achieved by retroviral vectors. For many malignancies, marking efficiencies are insufficient for useful information to be obtained. This problem may be overcome by the introduction of vectors that, unlike retroviruses, can stably integrate in cells that are not in cycle at the time of vector exposure. Other improvements will focus on the marker genes themselves, using marker elements that are simpler to track and will not produce any modification of the cells' behaviour. Finally, marker studies have proved safe so far, but follow-up of the treated patients continues.