Two studies have been performed to clarify the relationship between different markers of oxidative DNA damage commonly employed in molecular epidemiological studies. In the first, 8-Oxo-7,8-dihydroguanine (8-oxoGua) was induced in DNA of HeLa cells by treatment with different concentrations of photosensitizer Ro 19-8022 together with visible light. 8-OxoGua was estimated by the comet assay (alkaline single cell gel electrophoresis) with formamidopyrimidine DNA glycosylase and by HPLC with electrochemical detection. The dose-response curves indicate that the comet assay and HPLC are equally efficient at detecting induced damage. Background levels of 8-oxoGua in HeLa cells were 0.92 +/- 0.22 per 10(6) guanines by the comet assay and 2.09 +/- 0.13 per 10(6) guanines by HPLC. The second study was a small human trial, in which lymphocytes were collected for analysis of background levels of 8-oxoGua, as well as overnight and 24 h urine samples for measurement of excreted 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxodGuo) by ELISA. The mean level of 8-oxoGua in lymphocytes was determined as 1.33 +/- 0.21 per 10(6) guanines by the comet assay and 3.72 +/- 1.06 per 10(6) guanines by HPLC. A strong correlation was seen between overnight and 24 h urinary 8-oxodGuo (r = 0.93, P < 0.01). Overnight urinary 8-oxodGuo concentrations correlated with 8-oxoGua in lymphocytes measured by HPLC (r = 0.85, P < 0.05) or by the comet assay (r = 0.86, P < 0.05), although individual values from HPLC and the comet assay did not correlate with each other. It is reasonable to assess oxidative stress by any of these methods.