Reactive oxygen species (ROS) develop as a consequence of wounding, light stress and chemical imbalances but act also as signals in living cells. The integrity of cells is seriously endangered, if ROS cannot be controlled by scavenging molecules and other repair mechanisms of the cell. For studying ROS development and signalling under stress, a reliable indicator is needed. We have tested the ROS sensitive dye 5-(and-6) chloromethyl-2',7' dichlorodihydrofluorescein diacetate acetyl ester (CM-H(2)DCFDA) using onion bulb scale and leaf epidermis as well as Arabidopsis leaves and protoplasts. ROS were generated by several fundamentally different methods-externally applied hydrogen peroxide, heat shock, high light or wounding. Confocal microscopy and fluorescence quantification over time showed that the indicator responds in an additive and dose-dependent manner. The response to externally applied hydrogen peroxide followed saturation kinetics, consistent with a channel-mediated uptake of the stressor across the plasma membrane. An inherent problem of the tested indicator was the uneven uptake in tissues, as compared with protoplasts, making it difficult to discriminate an uneven indicator distribution from an uneven ROS distribution. However, in protoplasts and under carefully designed preparation conditions CM-H(2)DCFDA is a useful general ROS indicator. Subcellularly, the de-esterified probe localised to the cytosol, to mitochondria and to chloroplasts.