Earth's atmosphere is a disequilibrium mixture of oxidizable materials in the presence of large amounts of oxygen. The entropy reduction is brought about by living organisms using photochemical energy from the Sun. Almost all the oxygen in the atmosphere is derived from photosynthesis; inorganic photochemistry can produce only 10 −9 of the present atmospheric level at most. Oxygen is a critical component of our atmosphere, not so much because some living organisms employ respiration, but because O 2 is a precursor of atmospheric O 3. Between them, O 2 and O 3 filter out short-wavelength UV radiation from the Sun, and so permit the existence of life on dry land. The gases of the atmospheres of Venus, Earth and Mars have all been outgassed from within the planets after their formation. The most probable starting materials in the early palaeoatmospheres are CO 2, N 2 and H 2O, perhaps together with a little H 2 and CO. Photochemical transformation of CH 4 and NH 3 occurs in periods negligible on a geological time scale. The pressures, temperatures and composition of the atmospheres of Venus and Mars can be understood in terms of simple physical and photochemical processes. The behaviour of Earth is more complex, with changes in CO 2 concentration having kept planetary temperatures within a narrow band for billions of years, and the CO 2 having largely been replaced in the atmosphere by O 2. On Earth, most of the CO 2 orginally present seems to have become combined in carbonate rocks or dissolved in the oceans. Prebiological photochemistry involves mainly the photolysis of H 2O and CO 2. “Shadowing” and the need to remove H 2 from the atmosphere limit the amount of O 2 that can build up to very low levels. Ozone could not, therefore, have been an effective shield for the earliest life forms. Elemental sulphur, S 8, from the photolysis of SO 2 or H 2S, might be an alternative atmospheric UV filter. Photochemistry in an atmosphere containing H 2O and CO (or CO 2) can yield organic compounds that might ultimately form the precursors of living organisms: HCHO is especially important in this respect. Photolysis of CH 4 yields CH 2, and this species can participate in further organic chemistry. In addition, nitrogen atoms from high altitude photodissociation and photoionization of N 2 react with CH 2 to give HCN, so that HCN photochemistry might have played a role in the prebiotic atmosphere. Geological and fossil evidence can be used to infer atmospheric O 2 concentrations from the time at which life emerged 3.5×10 9 years ago or more. Photochemical atmospheric models can then be used to calculate probable atmospheric O 3 levels throughout geological time. A fascinating extension is to examine if the shielding potential of the amounts of O 3 present can be linked with identifiable evolutionary steps, and especially with the emergence of life from the protection of water to dry land.