The aim of this study was to compare three methods of detection of apoptotic cells: (1) the method based on elution of low molecular weight DNA from the ethanol fixed cells followed by cell staining with DAPI (diamidino-2–phenylindole) or propidium iodide as the DNA fluorochromes, (2) the method of in situ labeling of DNA strand breaks with biotinylated dUTP, utilizing exogenous terminal deoxyribonucleotide transferase, and (3) the method of analysis of DNA denaturation in situ using acridine orange to differentially stain denatured and doublestranded DNA sections following cell exposure to 0.1 M HCl. Cells of the human promyelocytic HL-60 line, treated in vitro with the DNA topoisomerase I inhibitor camptothecin, which selectively triggers apoptosis of S-phase cells, were chosen as a model. The method based on analysis of changes in DNA denaturability was the most sensitive in terms of detection of the earliest changes in chromatin of cells undergoing apoptosis; the increased sensitivity of DNA to denaturation in Sphase cells was measured as early as 100 min after addition of camptothecin. DNA cleavage, assayed either by the univariate measurement of DNA content following extraction of low molecular weight DNA, or by labeling DNA strand breaks with biotinylated dUTP, was detected in S-phase cells after 120 min incubation with camptothecin. The percentage of apoptotic cells at the late stage of apoptosis, the kinetics of cell transition to apoptosis, and kinetics of the loss of S phase cells were all essentially similar when measured by any method. All three methods can be used to estimate the cell cycle phase specificity of apoptosis, although the method based on DNA strand labeling with biotinylated dUTP by terminal deoxynucleotidyl transferase has the advantage of making it possible to estimate the cell cycle distribution of both the apoptotic and unaffected cell populations. The latter method also appears to be the most specific in terms of detection of apoptosis. © 1994 Wiley-Liss, Inc.