Hydrogen peroxide (H(2)O(2)) can be detected in freshly-voided human urine from healthy subjects and has been proposed as a "biomarker" of oxidative stress. This paper summarizes our studies to examine the extent to which urinary H(2)O(2) measurement fulfils the criteria for the "ideal biomarker". Levels of H(2)O(2), standardised for creatinine, varied widely between subjects. In most subjects, levels also varied considerably when measurements were made at different times and on different days. A reproducible increase in urinary H(2)O(2) was detected in all subjects examined after drinking coffee, a beverage rich in H(2)O(2). By contrast, green tea decreased urinary H(2)O(2) levels. We conclude that the H(2)O(2) in coffee is not excreted into urine. Instead, hydroxyhydroquinone from coffee is absorbed, excreted and oxidises in urine to produce H(2)O(2). No other confounders of urinary H(2)O(2) have been identified to date. Work is underway to compare H(2)O(2) levels with variations in other biomarkers of oxidative damage, to test the possibility that there are daily or other periodic variations in oxidative damage rates.