Superoxide (O2-) and hydrogen peroxide (H2O2) concentrations ranging from 87 to 1120 pmol L-1 and 5 to 107 nmol L-1, respectively, were measured in samples of surface water from the Great Barrier Reef (GBR) lagoon in the absence of photochemistry. Nonphotochemical, particle-associated net production rates of O2- ranging from 1 to 16 pmol L-1 s-1 were also determined and calculated to be similar in magnitude to the likely abiotic photochemical O2- production rates in GBR surface waters. Manipulative experiments using 0.22-µm filtration and addition of biological inhibitors demonstrated that the majority of this particle-associated production was biological and likely driven by photosynthetic organisms. Pseudo-first-order O2- decay rate constants were very low at O2- concentrations < 1 nmol L-1 (values in filtered samples ranged from 0.7 to 4.3 x 10-2 s-1) but increased with increasing O2- concentration toward a value of ~ 0.2 s-1 at O2- concentrations > 10 nmol L-1. This was thought to occur because reduced forms of metals such as iron and copper, or redox-active organic moieties, preferentially react with O2 rather than O2- at low O2- concentrations, thereby inhibiting catalyzed O2- disproportionation. This notion was supported by the observation that addition of superoxide dismutase dramatically increased rates of H2O2 production in samples. We suggest that, under these conditions, O2- can maintain a biologically useful reducing microenvironment around cells without resulting in significant accumulation of potentially harmful H2O2