Preservatives are added to many final products, such as detergents, cosmetics, pharmaceuticals and vaccines. We conducted an in vitro investigation of the apoptosis- and necrosis-inducing potential of brief applications (10 min) of four common preservatives: ethylene glycol monophenyl ether, 2-phenoxyethanol (EGPE), imidazolidinyl urea (IMU), a mixture of 5-chloro-2-methyl-4-isothiazolin-3-one and 2-methyl-4-isothiazolin-3-one (CMI/MI), and 1,2-pentanediol, a "preservative-non-preservative" best known as pentylene glycol. Using HL60 cells, we monitored the kinetics of cell toxicity with the MTT test and analysed extranuclear end points of apoptosis, i.e. phosphatidylserine exposure and nuclear fragmentation. Preservative treatment resulted in a dose-dependent decrease of cell viability. The mode of cell death was dose-dependent: necrosis occurred at high concentrations while apoptosis, shown by DNA laddering, DNA sub-diploid peak and caspase-3 activation, occurred at lower concentrations 0-24hr after exposure to a single dose: CMI/MI induced apoptosis at low concentrations (0.001-0.01%) and necrosis at high concentrations (0.5-0.1%); IMU and EGPE required higher concentrations to induce apoptosis (IMU 0.01-0.1% and EGPE 0.01-0.5%) or necrosis (IMU 0.5-1% and EGPE only at 1%). PG induced apoptosis only at 5%. Externalization of PS, a hallmark of apoptosis, occurred early in HL60 treated with low concentrations of CMI/MI and EGPE and was concomitant with the subdiploid peak in HL60 treated with PG. However, it did not occur in HL60 treated with IMU. In conclusion, at appropriate concentrations, each of the four preservatives modulates the apoptotic machinery by a caspase-dependent mechanism. Thus, apoptosis could be a good parameter to evaluate the cytoxicity of these chemical compounds.