Abstract In a series of recent papers, Kiess and collaborators have announced the identification of nitrogen peroxide on Mars and have attempted to interpret a wide range of Martian phenomena in terms of nitrogen oxides. Analyses of observations at a variety of wavelengths in the present paper, and previously by Sinton, Spinrad and Moroz, place a firm upper limit on the Martian NO 2 abundance of 1 mm-atm. Nitrogen dioxide is a highly photolabile gas and will be photodissociated by visible and ultraviolet radiation on Mars. The photochemical equilibria of nitrogen oxides on Mars have been computed from the observational upper limits on the NO and O 2 abundances. The same procedure gives consistent results for the Earth, at locales free from urban pollution. The corresponding theoretical upper limit to the abundance of NO 2 on Mars is 1 mm-atm, when reactions with water are neglected. When reactions with water are considered, the NO 2 abundance is further diminished. These low abundances are inadequate to account for the Martian observables discussed by Kiess et al. The one Martian phenomenon in which nitrogen peroxide may play a role is the blue haze, where 1 mm-atm may help to explain not only the general blue and violet opacity, but also the dependence of the opacity on time and position. The required abundance of NO must then be >50 cm-atm. Kiess et al . have expressed the opinion that the quantity of nitrogen peroxide on Mars is so large as to exclude indigenous life. It is of interest to note that a typical value for the abundance of NO 2 over the city of Los Angeles exceeds the corresponding value for Mars.