Hydrogen peroxide (H₂O₂) is widely distributed in surface waters where the primary photochemical formation pathway involves the interaction between dissolved organic carbon (DOC) and ultraviolet radiation (UVR). In laboratory studies using iron-rich water from Yellowstone's Chocolate Pots spring, H₂O₂ formation depended on sample treatment (unfiltered, <0.2 μm filtered, autoclaved) prior to irradiation, suggesting several formation pathways. Similar H₂O₂ formation in filtered and unfiltered water indicates that it is primarily soluble material that is responsible for H₂O₂ formation. H₂O₂ formation with soluble material probably includes only photochemical reactions with DOC and/or metals. Greater H₂O₂ formation in unfiltered and filtered water than in autoclaved water suggests that the agent(s) involved in H₂O₂ formation is (are) not stable at high temperatures and pressures and degrade to nonphotoreactive species. Such unstable agents may include DOC and/or dissolved complexes of iron or other metals. UVR absorbance occurs across the UV spectrum and, though slightly greater in the UVA range (320–400 nm), is similar to that of other surface waters. Increased UVR absorbance after autoclaving suggested degradation or alteration of some components, which in turn affected H₂O₂ formation. The spectral region used for irradiation affected net formation and yield. H₂O₂ formation in water irradiated with UVA radiation was 2.5–3 times that formed in water irradiated with UVB radiation (280–320 nm) in experiments using artificial light sources. Apparent quantum yields comparable to those reported by others could not be calculated because the instrumental designs are not the same. However, approximate quantum yields were calculated for these experiments but should be viewed with caution. Quantum yields were higher in these experiments (0.0040 mol H₂O₂ per mol photon at 310 nm and 0.0012 mol H₂O₂ per mol photon at 350 nm) than values reported by other researchers (<0.0007 mol H₂O₂ per mol photon at 300 nm and <0.0005 H₂O₂ per mol photon at 340 nm; [Scully, N. M., D. R. S. Lean, D. J. McQueen and W. J. Cooper (1996) Limnol. Oceanogr. 41, 540–548]). In natural solar source experiments, H₂O₂ formation was greater in experiments with UVA and photosynthetically active radiation (PAR; 400–700 nm) than with PAR alone or with UVB, UVA and PAR. However, H₂O₂ capacity (nM H₂O₂ W⁻¹ h⁻¹ m²) was greatest with UVB radiation and lowest with PAR radiation. Source regions could not be studied separately. Dark decay of H₂O₂ occurred via two mechanisms. The main mechanism responsible for H₂O₂ decay involved particulate matter (probably microorganisms), whereas a secondary mechanism involved soluble matter (i.e. DOC, metal ions and other dissolved species involved in Fenton reactions).