In the absence of a universal specific molecular tracer of apoptosis, structural DNA alterations provide the basis of labeling systems: double-strand fragmentation for TUNEL (terminal transferase-mediated dUTP nick end-labeling), denaturation for poly (A) in situ hybridization, immunogenicity of single strand DNA, all methods which imply limited specificity due to the unavoidable presence of DNA breaks in virtually all cells. Thus, TUNEL application has been restrained to a narrow spectrum of sample conditions which has limited, in particular, retrospective surveys and apoptotic nuclei-protein double labelings. In the apoptotic nucleus two main obstacles intervene between TUNEL reagents and their targets: DNA hypercondensation and proteins around DNA. The former increases in the course of apoptosis and both are worsened by crosslinking and precipitating fixatives. This point out that TUNEL is an ambitious approach whose target, apoptotic DNA breaks, is less accessible than breaks occurring in non-apoptotic less compacted DNA. However, TUNEL has an advantage: the far greater degree of apoptotic DNA fragmentation. How to obtain a frank differential staining between apoptotic and non-apoptotic DNA? It appears that the answer relies on the pretreatment step and not in modifying the TUNEL staining protocol, which is optimal. Adapted pretreatments are able to circumvent accessibility obstacles and to extend TUNEL applicability to the most demanding conditions, those of archived tissue samples and of TUNEL--protein double labelings.