We applied the nuclear DNA Diffusion Assay, described as an accurate tool to estimate apoptotic and necrotic cells [N.P. Singh, A simple method for accurate estimation of apoptotic cells, Exp. Cell Res. 256 (2000) 328-337] to tobacco root and leaf cells. In this assay, isolated nuclei are embedded in an agarose microgel on a microscope slide and low molecular-weight DNA fragments diffuse into the microgel. Exposure of the roots to hydrogen peroxide significantly increased the average nuclear area of isolated nuclei. After 4 and 24 h of recovery, all DNA damage was repaired. The data clearly demonstrate that the manifestation of diffused nuclei upon exposure to hydrogen peroxide is not the result of non-repairable apoptotic or necrotic DNA fragmentation, but represents repairable genotoxin-induced DNA damage. In contrast, treatment with the alkylating agent ethyl methanesulphonate (EMS) followed by 24 h of recovery produced a significant increase in the average nuclear area. The contribution of apoptosis to this increase cannot be excluded. Heat treatment of leaves at 50°C for 1-15 min leading to necrosis, and treatment of isolated nuclei with DNase-I, which digests DNA to nucleosome-sized fragments as during apoptosis, also led to a dose-dependent increase in the nuclear area. The use of different fluorochromes (ethidium bromide, DAPI or YOYO-1) for DNA staining yielded similar results in the DNA Diffusion Assay. As all types and sizes of diffused nuclei were observed after EMS and hydrogen peroxide treatments, we were unable to differentiate, on the basis of the structure of the nuclei, between apoptotic or necrotic DNA fragmentation and other types of genotoxin-induced DNA damage in plants. © 2005 Elsevier B.V. All rights reserved.