Most exoplanets detected so far have atmospheric temperatures significantly higher than 300 K. Often close to their star, they receive an intense UV photons flux that triggers important photodissociation processes. The temperature dependency of vacuum ultraviolet (VUV) absorption cross sections are poorly known, leading to an undefined uncertainty in atmospheric models. Similarly, data measured at low temperatures similar to those of the high atmosphere of Mars, Venus, and Titan are often lacking. Our aim is to quantify the temperature dependency of the VUV absorption cross sections of important molecules in planetary atmospheres. We want to provide high-resolution data at temperatures prevailing in these media, and a simple parameterisation of the absorption in order to simplify its use in photochemical models. This study focuses on carbon dioxide (CO₂). Methods. We performed experimental measurements of CO₂ absorption cross sections with synchrotron radiation for the wavelength range (115–200 nm). For longer wavelengths (195–230 nm), we used a deuterium lamp and a 1.5 m Jobin-Yvon spectrometer. We used these data in our one-dimensional (1D) thermo-photochemical model in order to study their impact on the predicted atmospheric compositions. The VUV absorption cross section of CO₂ increases with the temperature. The absorption we measured at 150 K seems to be close to the absorption of CO₂ in the fundamental ground state. The absorption cross section can be separated into two parts: a continuum and a fine structure superimposed on the continuum. The variation in the continuum of absorption can be represented by the sum of three Gaussian functions. Using data at high temperature in thermo-photochemical models significantly modifies the abundance and the photodissociation rates of many species in addition to CO₂, such as methane and ammonia. These deviations have an impact on synthetic transmission spectra, leading to variations of up to 5 ppm. Conclusions. We present a full set of high-resolution (∆λ = 0.03 nm) absorption cross sections of CO2 from 115 to 230 nm for temperatures ranging from 150 to 800 K. A parameterisation allows us to calculate the continuum of absorption in this wavelength range. Extrapolation at higher temperature has not been validated experimentally and therefore should be used with caution. Similar studies on other major species are necessary to improve our understanding of planetary atmospheres.