In recent years, the significance of apoptosis as a process in cell loss from normal tissue and tumours has been critically reviewed. In addition, the general lack of a correlation between radiation or drug-induced apoptosis and cell survival responses (using the clonogenic assay) in tumour cells has been demonstrated. Several different reasons have been discussed by other authors. It is the purpose of this review to argue that there are many different forms of cell death (terminal differentiation, micronucleation, mitotic catastrophe or multinucleation) that, like apoptosis, are regulated by the cell. In this context, apoptosis was the first cell death mechanism associated with active involvement of the cell (signal transduction). Furthermore, a large variety of different in vitro and a few in vivo models published so far show that the form of cell death can shift from, for example, mitotic catastrophe to apoptosis. The shift appears to be a general principle and depends on the cell model examined, the stressor type and the stressor intensity. These considerations help to explain the absence of a simple link between apoptosis and clonogenicity and suggest how to overcome that limitation, which has implications for the significance of apoptosis where the diagnosis and prognosis of cancer are concerned.